' U h- A 5 MSU RETURNING MATERIALS: P1ace in book drop to LJBRARJES remove this checkout from Jun-zyilll. your record. FINE§_wi11 be charged if book is returned after the date stamped below. “,3 1? M?- U 0 I 5 $2005 . .- M SOLAR TECHNOLOGIES AND THE SOFT PATH: AN EMPIRICAL EXAMINATION BY Dora G. Lodwick A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology 1987 Copyright by DORA JEAN LODWICK 1987 ABSTRACT SOLAR TECHNOLOGIES AND THE SOFT PATH: AN EMPIRICAL EXAMINATION BY Dora G. Lodwick A U.S. national probability sample of 2,023 traditional energy users and a purposive sample of 3,809 solar energy technology owners are compared to assess whether those who owned solar energy technologies in 1980 have more soft path preferences (SPP) (e.g. attitudes and behaviors compatible with soft energy path developments) than do those who did not own such technologies. It is suggested that SPP is a necessary but not sufficient condition for the social structural transformation of society to the soft energy path proposed by Amory Lovins. A soft path preferences scale is developed. The scale values for solar and nonsolar homeowners, for active and passive solar technology owners, and for those owning the technologies for four different time periods are compared. The nonsolar homeowners, passive technology owners, and those who owned the technologies from one to five years have the highest SPP scores. The greatest differences were found in the dimension of natural resources conservation. Partial least squares structural equation modeling is used Dora G. Lodwick to examine an extension and specification of Lovins' theory of soft energy path development. A model is created which focuses on the process of SPP development. Energy vulnerability needs, contextual resources, type of solar technologies, technological problems, and evaluation are the independent variables of the model. The hypothesis that the solar energy technologies have the strongest influence on the development of SPP was not supported except for the passive solar technologies owners. Contextual resources provided the strongest influence, although it was negative for the solar technology owners and positive for the nonowners. It is proposed that the symbolic nature of the solar technologies dominates the experiences of the owners. A bifurcation of the renewable energy base change process is suggested: (1) supportive of the hard energy path, changes driven by "energy as commodity" orientation of active solar systems owners and (2) changes more compatible with the energy path proposed by Lovins and driven by "energy as a natural resource to be conserved" perspective of passive solar technology owners. Policy, programmatic and research implications are explored. This work is dedicated to my parents, LOIDA AND FLOYD GRADY and my husband, WELDON A. LODWICK. They have believed in me. ACKNOWLEDGEMENTS This task was greatly enriched by the professional guidance provided by members of my Dissertation Committee: Dr. Craig Harris, Dr. Nan Johnson, Dr. Marvin Olsen, Dr. Allan Schmid, and Chair, Dr. Denton Morrison. Dr. Barbara Farhar-Pilgrim gave invaluable assistance in procuring the data tapes and in reviewing a previous draft of the manuscript. Dr. Frank Falk provided assistance with PLS. The Department of Sociology at Michigan State University provided support for acquiring the data tape from the Solar Energy Research Institute which sponsored the original study. Colleagues of the Department of Sociology at the University of Denver released me from some responsibilities during my writing and provided computer support. I am most grateful to Denton Morrison who has constantly responded "beyond the call of duty" to my fledgling efforts through the years. Bonnie Mbrrison has also been important. Weldon, Ryan and Stephen thanks for leaving me alone and carrying some of my home tasks. Maybe now I can play! TABLE OF CONTENTS LIST OF TABLES O O O O O O O O I O O O O O O O O O O O O I O O O O O O O O O O O ....... x LIST OF FIGMS O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O Xiii CHAPTERS I THE PROBLEM: OVERVIEW AND ORGANIZATION OF THE STUDY.00....O...0.00...00......OOOOOOOOOOOOOOOOO. 1 IntrOduction 0....COO...IOOCOOOOOOOOOOOOOOO00.... Overview of the Lovinses' Ideas ................. Organization of This Study ...................... \ocsw II APPROACHES TO TECHNOLOGY, ENERGY AND SOCIAL CHANGE 00.0.0.0...000......OOOOOOOOOOOOOOOOOOO00. 11 Classical Theories of Technology and Society .... 11 Energetic Theories of Society ................... 17 Values, Behavior and Technology ................. 23 Diffusion of Innovations Contributions .......... 30 The Diffusion of Energy Conservation Practices .. 33 The Diffusion of Solar Technologies ............. 36 III RESEARCH RELEVANT TO DIMENSIONS OF SOFT PATH PREFERENCES 0.0.0.0...OCCCOCIOOOOIOIOOOOCOOOOC.O. 42 The Social Context of the Lovins Proposal ....... 43 Changing Cultural Values .................. ...... 43 The “Energy Crisis" ............................. 45 The Solar Technology Choice ..................... 47 Research on the Dimensions of Soft Path Preferences ..................................... 48 Natural Resources Conservation Preferences ...... 48 Self-Reliance, Decentralization and Grassroots Democracy ....................................... 53 Personal Gains Preferences ...................... 63 vii IV VI VII Equity Impacts Social Diversity Impacts ........................ Transition Period Issues ........................ Other Criticisms of the Soft Path Theory ........ THEORETICAL SPECIFICATION AND MODEL: SPP Dmmmm 00......0000000000.000.000.00000. Lovins Theory of Soft Energy Path Social Change . “Vina. Ideas 0......OOOOOOOOOOOOOOOOOOOOOO00.... Suggested Additions ............................. Specification of Natural Resources Vulnerability Specification to Energy Vulnerability ........... Model Specifications ............................ RESEARCH AND SAMPLING DESIGN AND SAMPLE MOTERISTICS 0.0000000000000000...00......I.O. Research Design ................................. Data Gathering Processes ........................ Sampling Design ................................. Data Gathering Methods Obtaining the Data Tapes ........................ Instrument Design Samples' Characteristics ........................ Housing and Fuel Characteristics ................ Socioeconomic Characteristics ................... Other Demographic Characteristics ............... Regional Characteristics ........................ SPP DEVELOPMENT MODEL: OPERATIONALIZATION AND MEASUREMENT ISSUES 0.00.0.0...OOOOOOOOOOIOOOOOOO. Operationalizing the SPP Development Model ...... Energy Vulnerabilities Needs .................... Contextual Resources ............................ Technology Types ................................ Evaluation of Solar Systems ..................... Soft Path Preferences ........................... Measurement Issues C I C O O O O O O O O O I O O O O O O 0 O O O O O O O O O 0 ANALYSIS OOOIOOOOOCOOOOOOOOOO...0.00.00...0..O... I ntrOduct i on to PLS I I O O O O O O O O O O O O O O I O O O 0 O O O O O O O 0 Comparing PLS and ML ............................ Rationale for Selecting PLS Rather Than ML viii 76 77 77 82 84 90 98 104 104 106 106 109 110 112 114 114 120 126 130 132 132 134 138 147 150 152 157 168 169 174 Madeling OOOOOOOOOOOOOOOOOOOOOOOOOOOOOCOOOOOOOOOO PLS Specifications of the SPP Development Model . Measurement Model for Solar and Nonsolar Homeowners OOOCOIOOOOOOOOOOOOOOI0.00.00.000.00... Measurement Model for Solar Active and Passive Homeomers.‘OOOOCOOOOOOOOOOOIOOOOI0.0.000...0.... Specification of the Theoretical Model .......... VIII FINDINGS COCOOOOIOOOOCOOOOOIOOO0.0.0.000...00.... Soft Path Preferences Scale ..................... Spp Scales: Solar - Nonsolar Homeowners ......... Spp Scales: Passive - Active System Owners ...... Solar Homeowners - Effect of Length of Ownership Soft Path Preferences Development Model ......... Theore tical Model - Manifest Variables Ralationships OOOOOOOOOOOOOOOOCOOOOOOOO0.00...... Theore tical Medels: Solar and Nonsolar Homeowners 0..OOOOCOOOOOOOOOOOOOOOOO0.0.0.0....0. Theoretical Models: Active & Passive System owners COCOOOOOOOIOOOOOOOOOOOOOOOO0.0.0.000...OI. MadifiedMOdel OOOOOOOOOOCOOOOIOOOOOOOOOOOOOOO0.0 IX DISCUSSION AND CONCLUSIONS ...................... Theoretical Implications ........................ Theore tical Implications of the Soft Path Preferences Scale................................ Theoretical Implications of the Soft Path Preferences Model................................ Policy and Programmatic Implications............. Policy Implications ............................. Implications for Energy Programs ................ Limitations and Suggestions for Further Research. Limitations or the Study 0 O O O O O O O O O O O O O O O O O O O O O O 0 Future Research Implications .................... conCIuSions OOOOOOOOOOOO0.00000000000000000000000 ENDNOTES 0.0.00IOOOOIOOOOOOOOO0.00.00...0......IO..0... APPENDIX A: APPENDIX B: .APPENDIX C: .BIBLIOGRAPHY SELECTED DESCRIPTIONS OF THE SAMPLES ...... OPERATIONALIZING THE SPP MODEL............. MODELS' ix CHMflERISTICS OOOOOOIOOOOOOOOOOOO 175 178 179 180 181 185 185 185 189 192 198 199 207 214 222 230 231 231 240 246 246 249 252 253 260 263 265 267 281 303 320 TABLES 10 11 12 13 14 15 16 17 LIST OF TABLES SOLAR TECHNOLOGY ACQUISITION PROCESS ............ HOUSING CHARACTERISTICS AND PRIMARY HEATING FUEL. AVERAGE ENERGY COSTS ............................ HOUSEHOLD COMPOSITION ........................... SOCIOECONOMIC CHARACTERISTICS OF HOMEOWNERS ..... INCOME AND OCCUPATION - SOLAR ................. INCOME AND OCCUPATION - NONSOLAR ................ DEMOGRAPHIC CHARACTERISTICS OF HOMEOWNERS........ REGIONAL DISTRIBUTION OF HOMEOWNERS.............. SPP DEVELOPMENT MODEL INDICATORS ................ INDEXES OF THE SPP DEVELOPMENT MODEL ............ SPP SCALE WEIGHTS: SOLAR AND NONSOLAR ........... MODE SPECIFICATIONS FOR THE THEORETICAL MODELS .. SPP SCALE COMPARISON: SOLAR - NONSOLAR .......... SPP SCALE COMPARISON: ACTIVE - PASSIVE OWNERS.... SPP SCALE COMPARISON: LENGTH OF OWNERSHIP ....... BASELINE MODELS - MANIFEST VARIABLES: SOLAR AND NONSOLAR HOMEOWNERS ................... Page 94 115 117 119 121 123 124 127 131 133 159 165 182 186 191 194 201 18 19 20 21 22 23 24 25 26 27 28 29 3O 31 32 33 34 35 36 37 38 39 THEORETICAL MODELS - MANIFEST VARIABLES: SOLAR.... THEORETICAL MODELS' RELATIONSHIPS AND GOODNESS OF FIT: SOLAR AND NONSOLAR HOMEOWNERS .............. THEORETICAL MODELS' MANIFEST VARIABLES: ACTIVE AND PASSIVE TECHNOLOGY OWNERS.................... THEORETICAL MODELS' RELATIONSHIPS AND GOODNESS OF FIT: ACTIVE AND PASSIVE TECHNOLOGY OWNERS........ MODIFIED MODELS - MANIFEST VARIABLES: SOLAR, ACTIVE AND PASSIVE HOMEOWNERS ................... MODIFIED MODELS' RELATIONSHIPS AND GOODNESS OF FIT: SOLAR, ACTIVE AND PASSIVE HOMEOWNERS........ MOVING PLANS: SOLAR AND NONSOLAR HOMEOWNERS ..... PERCEPTIONS OF ”THE ENERGY CRISIS” .............. CONTEXTUAL RESOURCES: PERSONAL NETWORK SUPPORT FORSET O...OOOOOOOOOOOOOOOOOOOOO0.00.00.00.00..0 CONTEXTUAL RESOURCES: TECHNOLOGICAL SUPPORT ..... CONTEXTUAL RESOURCES: GOVERNMENT SUPPORT FOR..... SOLAR ENERGY TECHNOLOGIES TECH TYPE: SOLAR TECHNOLOGY OWNERSHIP LENGTH .... SOURCES OF PARTS AND MATERIALS FOR SELF MADE SYSTEMSOOCOOOOOOOOOOOOOOOOOOOOOOOOIOOOOOOOOOOO.I. PROBLEMS EXPERIENCED WITH SOLAR TECHNOLOGIES..... EVALUATION OF SOLAR TECHNOLOGIES ................ EVALUATION: INTENTIONS AND RECOMMENDATIONS....... SPP: SELF RELIANCE PREFERENCES .................. SPP: NATURAL RESOURCES CONSERVATION PREFERENCES.. SPP: ENERGY CONSERVATION ACTIVITIES ...... ....... SPP: PERSONAL GAINS PREFERENCES ................. PEARSON CORRELATIONS: SOLAR HOMEOWNERS........... PEARSON CORRELATIONS: NONSOLAR HOMEOWNERS........ xi 202 210 215 220 225 227 267 268 269 270 271 272 273 274 275 276 277 278 279 280 249 307 40 41 42 43 44 45 46 47 48 49 50 SPP SCALE WEIGHTS: ACTIVE AND PASSIVE TECWOWY OWNERS...oeeeeeeeoeeeeeseeeeeeeeeeesoe COMMUNITY PARTICIPATION: SOLAR AND NONSOLAR HOMOWERS...OOCCOIOOOCOOOOOOOOOCOCCO......0.0.0. UNOPERATIVE PERIODS OF SOLAR SYSTEMS............. THEORETICAL TRIMMED MODELS - MANIFEST VARIABLES: SOLAR AND NONSOLAR HOMEOWNERS ................... THEORETICAL TRIMMED MODELS - MANIFEST VARIABLES: ACTIVE AND PASSIVE TECHNOLOGY OWNERS............. THEORETICAL TRIMMED MODELS - GOODNESS OF FIT: SOLAR AND NONSOLAR HOMEOWNERS ................... THEORETICAL TRIMMED MODELS - GOODNESS OF FIT: ACTIVE AND PASSIVE TECHNOLOGY OWNERS............. MODIFIED TRIMMED MODELS - MANIFEST VARIABLES: SOLAR, ACTIVE AND PASSIVE TECHNOLOGY OWNERS ..... MODIFIED TRIMMED MODELS - MANIFEST VARIABLES: ACTIVE AND PASSIVE TECHNOLOGY OWNERS ............ MODIFIED TRIMMED MODELS - GOODNESS OF FIT: 80mHOMOWERS.OOOOOOOOOOOOOOOOOOICOO...0...00. MODIFIED TRIMMED MODELS - GOODNESS OF FIT: ACTIVE AND PASSIVE TECHNOLOGY OWNERS............. xii 309 310 311 312 313 314 315 316 317 318 319 LIST OF FIGURES FIGURES Page 1 LOVINS' VALUES AND TECHNOLOGICAL CHANGE MODEL ... 25 2 BASIC NATURAL RESOURCES VULNERABILITY............ 86 3 TRANSITION TO THE SOFT ENERGY PATH .............. 97 4 DEVELOPMENT OF SOFT PATH PREFERENCES ............ 99 5 SPP SCALE COMPARISON: SOLAR-NONSOLAR............ 186 6 SPP SCALE COMPARISON: ACTIVE-PASSIVE............ 191 7 TOTAL SPP SCORES COMPARISON: LENGTH OF OWNERSHIPOOOOOOOOOIO.......OOOOOOOO......OOOOOOOO 195 8 SOLAR THEORETICAL MODEL WITH PLS SPECIFICATIONS.. 208 9 NONSOLAR THEORETICAL MODEL WITH PLS SPECIFICATIONS CO.......COOOOOCOCCCOOOOOOO0...... 209 10 MODIFIED MODEL WITH PLS SPECIFICATIONS .......... 223 xiii CHAPTER I THE PROBLEM: OVERVIEW AND ORGANIZATION OF THE STUDY Introduction: Sociologists have long been concerned with the interaction of technology and society. This concern is scattered throughout the discipline, including the classical theorists such as Weber, Marx, Ogburn, Mannheim and others. Macro theories have examined the interaction of technology and society using either technology or society as the dependent variable. Many of the thinkers concerned with the Industrial Revolution have treated technology primarily as an independent variable restructuring society. Other theorists have examined technology as dependent. An important strand of the macro level tradition is captured by "energetic theories" (Rosa and Machlis, 1983), which portray technology as an intervening variable between physical energy and society. These theories have predominantly portrayed social change as a dependent variable. The second major approach to the study of technology and society has focused on the micro-level interactions. Studies 2 in the diffusion of innovation tradition and in the social psychology of values and behavior have been part of this approach. Micro theories have generally treated energy or technology as dependent variables affected by socially structured choices and values of individuals or households. Few studies have integrated the two levels of analysis looking at the intersection of technology, values, and the restructuring of society. In fact, Gaston (1980:496) has suggested that values have been neglected in the sociology of science and technology. This study empirically examines Amory Lovins' influential claim that the use of "soft" energy technology (SET) affects a package of attitudes and behaviors which will directly restructure society's energy base and indirectly its political and economic systems. "Soft" energy technologies are, according to Lovins, those that have ”soft" characteristics and impacts. They are: (1) natural resource conserving, (2) renewable resource using (e.g. based on renewable energy flows), (3) usually small in scale, (4) understandable (e.g. simple), (5) diverse, (6) matching energy quality to end-use needs, and (7) under the control of the end user (e.g. usually decentralized) (Lovins, 1977: 1978). I have called the package of values and behaviors soft path 3 preferences (SPP)1. This is based on Amory Lovins' work (1976, 1977, 1980). He has labelled the restructured society, which is based on SET's the "soft energy path (SEP)." This path is characterized as a complex, interacting set of mutually reinforcing, internally consistent features that together constitute an energy system that is, in effect, a sociotechnical system. (Morrison and Lodwick, 1981:365) Lovins et al. clarified their evolving perspective on the meaning of a "path" by saying that "The soft technologies form the base for an alternative policy known as the soft energy path" (Lovins et al., 1983:57). The path change assumes that a summation of micro-level household and community choices eventually translates into changes at the macro-level unless constraining barriers are erected preventing such a social structural transformation. Specifically, I examine the effect which the ownership of solar hot water heating and home heating and cooling systems has on household members' attitudes and behaviors. The central question I address is: Are the preferences expressed by households which own solar energy technologies more consistent with those needed for a social-structural transformation to the soft energy path than among those who do not own the technologies? I address this question by developing a theoretical schema 4 and model about the process of SPP development from the 2 Lovinses' and others' research. I then operationalize the model and test it with data from a national sample of households. In the following chapters, the present research is set in the context of broadly relevant social science thinking on technology and society, on energy and society, and the relevant social science empirical research on energy. The theoretical schema and model are discussed in detail in Chapter IV. The following brief overview of Lovins' ideas is a preliminary orientation to what will later be considered more thoroughly. Overview of the Lovinses' Ideas: Amory Lovins (1976, 1977, 1980) explores ways in which energy, through technologies, influences society. Although he has not been concerned with disentangling the precise causal relationships between energy, technology, social structure and values he basically posits the energy base of society as the independent variable affecting social structure (1977:153). His primary concern is how to counter the power of current socioeconomic and political structures which have developed what he claims is a destructive trajectory for the social system i.e. the hard energy path (HEP) through use of an inappropriate energy base. 5 Lovins favors changing the energy base of societies built on energy stock forms (e.g. oil, coal, gas, and especially uranium) to one based on flow forms (e.g. solar, wind, water) (Lovins, 1977:169: Morrison and Lodwick, 1981:367). If the technological and the accompanying sociopolitical structures of the energy base are changed, then there will be a change to the soft energy path which will entail significant social change (Lovins, 1977:54; Morrison and Lodwick, 1981). Individual choices made for technologies based on energy flows are grounded in personal values of “thrift, simplicity, diversity, neighborliness, humility and craftsmanship" (Lovins, 1977:57). Lovins claims choices based on these values will aggregate and, in turn, lead to changes in the political-economic system. The argument he makes is that the characteristics of technologies create opportunities for the expression of these values although the technological characteristics are not a sufficient condition for the development of soft path preferences. The previously sketched characteristics of SET's include the importance of socially organizing their implementation in a ”soft" fashion. This implementation process is noncoercively Jbased on user participation in making, operating and maintaining the technology (Morrison and Lodwick, 1981:366). The "soft" characteristics of the technologies as well as the "soft" implementation process are both necessary and sufficient conditions for creating "soft" social impacts e.g. a move to SEP. If both conditions are not present, then the technology only has the potential for being a soft technology (Lovins, 1977:42: Morrison and Lodwick, 1981:369). Lovins claims that rapid value changes are currently occurring in the U.S. which supplement the sociotechnical changes (1977:36) and which in turn are further reinforced by the use of the new energy technologies. There are five major dimensions of preferences that define a conceptually integrated package of these values: (1) Central to this package is the importance of decentralizing the primary economic and social institutions of the nation. This is captured in the notion of gel; reliance preferences. Values, attitudes and behaviors which support local (e.g. community and neighborhood) political and economic activities as opposed to national or international activities are an integral part of SPP i.e. grassroots democracy (Lovins et al., 1983; Lovins and 7 Lovins, 1982: Kinsley, 1984; Morrison and Lodwick, 1981). (2) The personal gain preference is also an important component of SPP. Lovins relies on the "free market" concept suggesting that if individuals make choices based on their individual economic and noneconomic benefits, these choices will help drive other aspects of SEP(e.g. resources conservation, equity). This has not occurred in the past because the market has been distorted by government subsidies to organizations driving the development of the hard path e.g. oil companies, centralized utility companies, etc. (Lovins, 1977: 1978: 1980: Lovins et al., 1983: Morrison and Lodwick, 1981). (3) The third dimension, natural resources conservation preferences, embraces an increased recognition of the finiteness of the physical world and of the need to conserve many types of natural resources so that other generations and nations have sufficient resources to experience a better quality of life than if the resources are not conserved (Lovins, 1976: 1977, 1980: Lovins and Lovins, 1982; Lovins et al.: 1983). (4) The fourth preference which is part of SPP is a concern for equity. Lovins suggests that a more equitable access to 8 and distribution of energy resources is an important outcome of the implementation of the soft energy path for individuals, communities, social classes, nations and generations. This access and assumed control will occurr with the implementation of soft technologies in a decentralized manner. It will create a greater surplus of energy resources than if the soft path is not implemented (Lovins, 1977, 1978; Lovins and Lovins, 1982: Lovins et al., 19833M0rrison and Lodwick,l981). (5) Social diversity preference is also a result of the implementation of energy technologies based on end-use, non- coerciveness, economic self-interest and grassroots democracy (Lovins, 1976, 1977; Lovins and Lovins, 1982; Lovins et al., 1983; Morrison and Lodwick, 1981). These last two are major social impacts of the development of the soft path. They will occur regardless of the values involved in implementation simply as a result of the use of diverse technologies and increased availability of energy. However preferences for such impacts are compatible with and theoretically should speed the development of the soft path. Thus Amory and Hunter Lovins are part of the theoretical tradition defining technology as the translator of the impacts of energy onto the social system. The amount and type of energy used in a society are the primary forces behind social structure and social change. While Amory 9 Lovins3 is not a social scientist, but rather a physicist turned energy activist, his ideas fit into a sociological tradition which examines the interaction of technology and society. The Lovinses' ideas have been socially and social scientifically important and influential. While somewhat general and abstract, they are sufficiently specific to lend themselves to further theoretical development, operationalization and testing. They have been basically unexamined theoretically and empirically, thus inviting further study. Organization of This Study: In Chapter II, I review what social scientists have previously analyzed about the relationship of energy, technology and social change. I begin with a brief look at the classical theorists, then focus on energy and society analysts. In Chapter III, details of more recent studies bearing directly on the issues at hand are discussed. Social scientific criticisms of the Lovinses' ideas are also reviewed. Using the criticisms and social scientists' research as a foundation, I extend the Lovinses' theory of social change, specifying how soft path preferences are developed at the household level. This specification is 10 then translated into a testable model of the process in Chapter IV. Chapter V includes the research design and a description of the sampling and data gathering processes. Descriptive information about the samples is also presented. In Chapter VI, the model's variables are operationalized and measurement issues are presented. In Chapter VII, I discuss the partial least squares analysis technique and specify how I used it to model soft path preferences development. Chapter VIII presents the findings of the study. Finally, in Chapter Ix, I discuss my conclusions, policy implications and future research. CHAPTER II APPROACHES TO TECHNOLOGY, ENERGY AND SOCIAL CHANGE I begin by discussing the classical theories about the relationship between technology and society which were fomented by the Industrial Revolution in Europe and the 0.8. The second theoretical tradition examined is of the energetic theories. A discussion of research about the interrelationships between technology, values and behavior then links research on technology, energy, and social change to the diffusion of innovations research traditions. The diffusion of energy conservation practices, and of solar energy technologies, are two very relevant areas of research for notions of soft path changes. Classical Theories of Technology and Society: Sociology has examined the interrelationship of technology and society from the moment of its birth as part of the social sciences. Bernard de Mandevilles first expressed this concern as he watched the birth of the Industrial Revolution in England (Weinstein, 1982:12). The classical theorists of technology and society are primarily concerned with how technology is used by elites to structure the social system for the purposes of industrial production to serve the bourgeois and control the political 11 12 processes of society. The gulf between elites and the general population (e.g. the proletariat) is a major concern of these thinkers. The interplay of the values of the dominant social class and technological developments is thus part of the contributions of this literature. Karl Marx and Friedrick Engels' writings on the impacts of the new industrial as well as of the ancient and prehistoric technologies on the organization of the economy, the state, the family, and culture have identified them as one of the earliest "technology assessors". They perceived technology as a force of production and examined the struggle between the means and relations of production which assured conflicting and uneven developments. The material base creates a social stratification system based on these conflicts (Weinstein, 1982). Max Weber was also concerned with technology although his influence has been more strongly felt through his focus on bureaucracy, organization and method. This became dominant in American sociology, taking over from the emphasis on social problems and concerns with the impacts of automation on social relations. There were, however, several exceptions to the dominant Weberian emphasis. Thorstein Veblen (1857-1929) and William F. Ogburn (1886-1959) both stressed how "values and social "P b. Ln. M‘- a... any, 'I'El‘ V... “on. G"-' 13 relations are systematically affected by - and affect - technological innovation" (Weinstein, 1982:44). Both in Veblen and in Ogburn's work, the emphasis on technology as a human, culturally-bound system with its potential for good or for evil became apparent. This is a theme which has appeared repeatedly in sociological studies of technology and society. Thorstein Veblen reasoned that engineers (e.g."technocrats") were constrained by the profit motive of the business class, thus preventing the freeing of technology to guide the evolution of society. His interest in the way cultural values, as expressed in the behavior of the social classes, affected technological developments, was examined through a comparison of England and Germany's experience with the technology of the Industrial Revolution (1939). German frugality, rather than ostentatious class consumption, created a more productive social context for the industrial revolution. To Veblen, technocrats were some of the most creative people of society who were stifled by their cultural milieu (Weinstein, 1982:49). They were not, therefore, able to positively guide societal change. William Fielding Ogburn explored many facets of the relationships of social change and technology, producing the first recognizable social impact assessment (see for 14 example, Ogburn, 1938). He also agreed that social groups, e.g. elites, used technology "to mask real interests" (Weinstein, 1982:56). Ogburn made an important contribution to legitimating the study of technology within sociology and in developing the notion of "cultural lag" (1922). But his emphasis on the independence and strength of technology in creating social change and his oversimplification of its effects on social life, have tended to label him a technological determinist (Weinstein, 1982:59). Karl Mannheim (1950) focused on the developments of what he termed Big Science in the late 1930's and early 1940's, a a trend which increased in speed during the post WWII period (Price, 1963). He took the position, that technological society is increasingly shaped by elite technicians and planners unresponsive to and insulated from the effects of the free market, public opinion, or egalitarian ideals. The increased specialization of the technological delivery system has made the rule by elites and the lack of public participation in the development of those systems inevitable. He was concerned with how the technological delivery system has replaced the laissez faire norms of the marketplace . Mannheim's work suggested that social scientists could help «develop a centralized democratic society which planned for freedom, for social justice, and for cultural standards. He 15 was a precursor of Lovins in calling for planning without regimentation (Mannheim, 1950;29). He proposed that value changes (e.g. "moral and religious awakening") needed to occur along with technological change to transform society. Some of the new values were to recapture pride of craftsmanship (Mannheim, 1950:223) and attendant responsibility. This is a theme repeated in Lovins' benefits of the soft path (Lovins, 1977: Morrison and Lodwick, 1981). The interaction of the "power elites" with engineers and technology in societal development processes was one focus of C. Wright Mills (1959, 1963). He was interested in the use of the sociological imagination to free people, a theme compatible with Lovins. The need for freeing the populace from elite technological dominance was developed further by members of the Frankfurt Institute for Social Research, established in Germany in 1923. Technology, in combination with economic and Psychological forces, is perceived as the key package of Variables shaping modern society and in turn being shaped by the society. This technology "has contributed significantly to the general dehumanization and trivialization of our lives, thoughts and aspirations" (Weinstein, 1982:103: Marcuse, 1968). 16 More recently, Jurgen Habermas (1970) has called for consciously mediating technical developments and life in major industrial societies (Habermas, 1970;60). Habermas suggests that the production relation legitimates itself and adapts political relations to the economic subsystem. Science and technology have become a legitimating force as they are perceived to serve economic development. Habermas claims that a new "conflict zone" is replacing class antagonisms "in the public sphere administered through the mass media" where the questioning of the nation's ”technocratic background ideology" is depoliticized (Habermas, 1970:120). This conflict zone has been entered by the Lovinses. They have pointed to the technocratic ideology and its effects, suggesting a substitute. The themes expressed by the classical theorists are consistent with Lovins' theory of technologically mediated social change. They include: (1) issues of technology control, (2) conflicts engendered by ownership of technology, (3) isolation of the controllers from the users of the technologies, (4) the importance of planning for change, and (5) the interaction of technology and values. The Lovinses have stressed the negative effects of the 17 dominance of the technocratic elite through the hard energy path. They have called for technical decisions on the part of households and local communities for technologies which are more consistent with prevalent values of the nonelites (Lovins, 1977: Morrison and Lodwick, 1981). Although the Lovinses' ideas are constrained by scientific and technical ideology supporting economic development, they have proposed creating social change by modifying the basic energy technology type of society (Kinsley, 1984; Lovins and Lovins, 1982: Lovins et al., 1983). Energetic Theories of Society: The second major theoretical tradition informing this research project in more specific ways than the classical tradition discussed above, examines the relationship between energy and society. In this tradition energy is the primary force driving technology. Energetic theorists often conclude that energy drives all of the facets of society. They also show that their concern has traditional roots. For example, in 1862, Herbert Spencer wrote: Whatever takes place in society results either from the undirected physical energies around, from these energies as directed by men, or from the energies of men themselves. (quoted in Carniero, 1967:xxxv by Rosa and Machlis, 1983:1) These theorists further examine the interaction of values with the way which energy is captured by certain types of 18 technologies to create social change. The "energy crisis" of the 1970's called attention to some of the implications of energy's permeation through the social system creating ”system vulnerability" (Schnaiberg, 1983). Lewis Mumford (1934, 1967) was one of the earliest social thinkers to specifically link technology, energy, and social values. He argued that knowledge of the amount of energy used by a technology was not sufficient to predict its use. The "culture that was ready to use" the technology also had to be examined (Mumford, 1934:4). He suggested that cultural development usually occurred before and after the use of a physical technology. This included values and social organization. To develop this thesis, he examined the interactions of values, technology, and energy during different periods of civilization. He identified coal and electricity as important sources of power helping to develop new civilizations. The coal based eighteenth century was described as a period when Mankind behaved like a drunken heir on a spree. And the damage to form and civilization through the prevalence of these new habits of disorderly exploitation and wasteful expenditure remained, whether or not the source of energy itself disappeared. The psychological results of carboniferous capitalism - the lowered morale, the expectation of getting something for nothing, the disregard for a balanced mode of production and consumption, the habituation to l9 wreckage and debris as part of the normal human environment - all these results were plainly mischievous. (Mnmford, 1934:158) He also critiqued classical economics for not giving adequate attention to the energy base of economic activity. For once energy is converted by the technology, it "runs down hill, in gathering and shaping the raw materials, in transporting supplies and products, and in the process of consumption itself" (1934:378). Thus he predated the writings of Georgescu-Roegen (1971), Odum (1971), and Rifkin (1980). Mumford puts it succinctly when he says: The real significance of the machines, socially speaking, does not consist either in the multiplication of goods or the multiplication of wants, real or illusory. Its significance lies in the gains of energy through increased conversion, through efficient production, through balanced consumptions, and through socialized creation. The test of economic success does not, therefore, lie in the industrial process alone, and it cannot be measured by the amount of horsepower converted or by the amount commanded by the individiaul user: for the important factors here are not quantities but ratios: ratios of mechanical effort to social and cultural results. (Mumford, 1934:378) In his examination of the evolution of humanity, Mumford (1967) stressed the numerous ”democratic technologies" of daily living (baskets, pots, barns, etc.). Many of these *were created by women (1967:141). They coexisted with an 20 authoritarian technology (e.g. centrally directed by dominant minorities) which was based on large scale social organization of people, the "megamachine" (Mumford, 1967:189). Almost from beginning of civilization, we can now see, two disparate technologies have existed side by side: one 'democratic' and dispersed, the other totalitarian and centralized. The 'democratic' mode, based on small-scale handicraft operations, was kept alive in a multitude of little villages, in partnership with farming and herding, though spreading into the growing country towns and finally lured into the cities. (Mumford, 1967:236) This theme of the two technologies foreshadowed the struggle Lovins claims exists today between hard technologies and soft technologies, which are socioculturally incompatible (1977). Anthropologist Leslie White closely examined the relationship between energy and social progress. He determined that the degree of progress (e.g. economic development) was based on the amount of energy harnessed. It is the relationship between technology, its efficiency, and energy that directly and indirectly affects change in the culture and the social organization of societies. Societies evolve through finding new ways of harnessing and concentrating solar energy for culture building (White, 1959). 21 This "culture building" serves the elites, claims Richard (1975). Social power evolves and becomes more concentrated as increased energy is generated and used by society. Energy is thus a driving force in social stratification and change carried by "mentalistic" (e.g. values carried as information) systems. As Rosa and Machlis report, these theorists "argued that societal change and progress were directed by the amount of energy harnessed" which in turn was determined by the technology of the society (1983:11). These thinkers, however, had very little concern with the limiting aspects of the Second Law of Thermodynamics. Fred Cottrell (1955) focused on the limiting nature of energy. He suggested that the change from a low-energy society (e.g. agricultural) to high-energy society (e.g. industrial) depended on generating energy surplus which was reinvested into technology of high-energy converters. His position is that the essential requirement for this conversion is a ”system of values that promotes the creation and reinvestment of energy surplus" (Rosa and Machlis, 1983:14). The investment of energy into high energy developments concerns Georgescu-Roegen (1975). He claims that society 22 will have to rely on energy stocks (e.g. coal, oil, gas) for such developments because energy flows (e.g. solar, wind, water) are too limiting in their dissipated forms. Since energy stocks are limited, having been created over millions of years, overconsuming behavior of the present population will use up the energy base of society. He does not think that resources are infinitely substitutable or that technologies can be continually created to compensate for the decreased quality of the energy. In this, he was strongly attacking a classical economic perspective as have others (Rifkin, 1980: Thurow, 1980). Contradicting Lovins' proposal of creating a more resilient energy system through individual efforts, Howard Odum (1971) warned about the dangers of giving individuals control over their energy base. Social coordination is necessary to prevent individuals from succumbing to "environmental whimsy." He developed a systems approach to the influence of energy on society, suggesting that energy flows influence economics, politics and religion. Amory Lovins' ideas are not tightly embedded in the social science literature, but nevertheless the basic conceptual framework of Lovins flows directly from the energetic theorists concern with the way in which the societal energy Ibase affects the social structures of societies. Some of the themes of this tradition which inform his work are: 23 (l) the importance of value complexes (e.g. culture) as a context and result of technological implementation, (2) the existence of dual technological streams, (3) the limiting nature of energy stock, and (4) the influence of the energy base on social structures. Lovins concern with equity is reflected in his desire that energy surplus created by an energy flow base be captured by the populace rather than social elites. He proposed the marketplace as the main mechanism for distributing technologies to change the energy base of society. This will help counter the historical concentration of energy, claims Lovins. Values, Behavior and Technology: The theories examined so far raise questions about the influence of culture and values as well as of whgge values or culture dominate technologically driven social change. Mannheim (1950) called for concomitant value and technological changes. He wanted technology to 22; be as closely controlled by insulated technocrats. C. Wright Mills, Marcuse, Adams, and Habermas fundamentally agree with his concerns about elites' control. Lovins is closest to Mumford in identifying two paths which are the results of implementing two different technologies - 24 the authoritarian (energy stock based technologies, elite controlled) or the democratic (energy flow based technologies, controlled by end users). Lovins and Mumford are also similar in suggesting that cultural changes (e.g. value changes) usually occur before the implementation of types of technologies as well as after their implementation. The question of whether technological change occurs in response to supportive cultural values or whether it changes values has a long and distinguished career. Robert Merton (1970) continued the tradition as he traced the evolution of the Industrial Revolution in seventeenth century England to cultural complexes such as Puritanism. On the other hand, Gouldner and Peterson (1962) examined 71 societies of the Human Relations Area Files to determine what were the most critical elements in social change. After conducting factor analysis of various societal characteristics, the authors concluded that the dominant factor was technology and the second one, values. Their conclusion was "technology influences the normative" (Gouldner and Peterson, 1962:xv). However, echoing Pitirim Sorokin (1937-1941), they also suggest that the dominance of technology may simply reflect a transitional historical period. The logical relationships between technology and values was closely traced by Emmanuel Mesthene (1970). He suggested 25 that technology leads to value changes both directly and indirectly. Directly, technology "appears to lead to value change either by bringing some previously unattainable goal within the realm of choice, or by making some values easier to implement than in the past, that is, by changing the costs associated with realizing them" (Mesthene, 1970:50). Indirectly, technology changes values through "the mediation of some more general social or cultural changes produced by technology" (1970:54). The effects of technology on values and social change within Lovins' work was traced by Schnaiberg (1983) as indicated in Figure 1. New Knowledge Socioeconomic of System 1; Value Change._____, Use of Energy Vulnerability (SEP) Technological Change/’////~ Fig. 1: LOVINS' VALUES AND TECHNOLOGICAL CHANGE MODEL (Schnaiberg, 1983:219) Social change therefore occurs directly as a result of value changes 95 as a result of straight technical fixes. I think it is truer to the Lovins' perspective, however, to also include an arrow to show the interactive relationship of technological and value changes. 26 As presented in Figure l, Schnaiberg's interpretation of Lovins is that new knowledge of system vulnerability is sufficient for technological or for value changes. This, however, is a misrepresentation of Lovins' argument. According to Lovins, it is not simply the new knowledge of system vulnerability that is the driving variable motivating individuals to change. Lovins proposes that the alternative energy base will be implemented because the renewable technologies provide a better expression of nonelite values than do the nonrenewable energy technologies. In his early work, Lovins emphasized the personal noneconomic and economic gains to be obtained from the renewable technologies. In later publications, the Lovinses have stressed community-wide gains such as using energy technologies as economic development strategies (Lovins and Lovins, 1982; Lovins et al., 1983). In suggesting that technologies are implemented when they express value preferences, Lovins is supported by Hornick and Enk's close examination of the interaction of values and technologies. They argue that "new technologies can play an important role in retargeting or implementation of our values by giving us new means to obtain our goals" (Hornick and Enk, 1980:85). Yet it is often true that there is not a high correspondence 'nhnsn “Hun U ‘ ems: ' I h L Iyu 3e“ ‘ ‘ 27 between expressed values and actual behavior (Schuman and Johnson, 1976: Hornick and Enk, 1980). This issue has been explored repeatedly in examining whether attitudes cause behavior or vice-versa. Bentler and Speckart (1981) examined the issue and concluded that one needs to examine specific domains to reach appropriate conclusions. They found that in three out of four situations which they examined - in dating, studying and exercise - attitudes preceded behaviors. However, behaviors preceded attitudes in one "studying" instance. They suggest that possibly behaviors influence attitudes when "internal cues are weak or ambiguous" (Bentler and Speckart, 1981:236). Once a behavior has occurred, it tends to create a more favorable attitude to more of that kind of behavior. They underscore, however, the critical effect of the domain on the interaction of behaviors, intentions and attitudes. Stern and Aronson (1984) pursue the idea that the use of technologies may change values at the individual level. They argue that people tend to rationalize choices they have made in a difficult decision situation. The more the commitment in cost, effort, and irrevocability, the stronger and more permanent the effect. Therefore, once a person makes a commitment in a direction, such as by using solar energy technologies, that person is more likely to 28 make a further large commitment than someone who is uninvolved (Stern and Aronson, 1984:69). This cognitive dissonance argument can be made for the direct influence of solar energy technologies in further developing behaviors and attitudes supportive of soft path changes. This is especially true if the implementation of the technologies is perceived as a commitment to natural resources conservation, self-reliance, greater equity, and social diversity - the dimensions of soft path preferences. Solar energy technologies have been identified as strongly symbolic of benign, clean living (Barbour et al., 1982: Hornick and Enk, 1980). Even though homeowners may not have thought of the values associated with the soft path when they acquired solar energy systems, the technology may serve as an instrument to further associate them with people and ideologies supportive of soft path preferences. However, energy is invisible to households, argue Stern and Aronson (1984). Numerous studies indicate that householders do not know their actual energy expenditures, the cost of solar technologies, or even how their technologies are operating (Unseld and Crews, 1980: Farhar et al., 1980: Farhar-Pilgrim and Unseld, 1982: Eastman, 1982). Energy's visibility is increased when people consider the use of solar te er. :3}; c Rea“ .. a U a o A: and bah Rite“ “eels: Q 3560 1; Va 29 solar technologies or when users are actually committed to energy conservation. Energy conservation may be the first behavioral commitment to the soft energy path and the use of solar energy technologies, the second step. Keating et al. (1982) explored the interaction of attitudes and behavioral intention using the Fishbein4 and diffusion models to predict solar adoption. The variables used in this prediction were: (1) environmental concerns, (2) perceptions of the seriousness of the energy situation, (3) number of contacts with homeowners having solar equipment, (4) voluntary simplicity lifetsyle behaviors, and (5) socioeconomic variables of income, education and occupation. They found that the most important predictor was attitudes toward solar technologies, followed by positive economic perceptions, negative economic perceptions and educational attainment. They suggest that the normative orientation of others was not important at this early stage of solar adoption when social expectations may not yet be formed (Keating et al., 1982). While Keating et al. examined individual level adoption of solar technologies, Gilmer focused on how technology influences institutional developments. He argued that: (1) technology constrains but does not determine institutional possibilities, and (2) some technologies are more easily managed than others (Gilmer, l980:3). This opinion is 30 echoed by others who suggest that technology creates one of the "outer frameworks of constraints on individual lifestyles" (Hornick and Enk, l980:4: Barbour et al., 1982). DIFFUSION OF INNOVATIONS CONTRIBUTIONS Two areas of diffusion research are especially relevant to the development of the soft path: (1) studies about the adoption of energy conservation practices, and (2) studies about the adoption of solar energy technologies. Diffusion researchers have examined the interaction of technology, values and behavior using a microsociological approach. This tradition reverses the causal order developed by macrosociologists by focusing primarily on the ways in which values, social class, lifestyle, etc. of individuals and households affect energy use (Rogers, 1983). In these studies, energy is a dependent variables (Rosa and Machlis, 1983:37). A tremendous amount of research has developed out of the diffusion of innovations tradition, making the steps involved in the adoption of different technologies very clear. Rogers presents the innovation-decision process model as including prior conditions (a) previous practice, (b) felt needs/problems, (c) norms of the social systems which provide the context for the five stages of adoption which 31 are connected through communication channels: (1) knowledge (characteristics of the decision-making unit such as socioeconomic characteristics, personality variables, and communication behavior), (2) persuasion (influenced by the perceived characteristics of the innovation regarding relative advantage, compatibility, complexity, triability, and observability), (3) decision (adoption or rejection), (4) implementation, and finally (5) confirmation of the adoption with continuance of the use of the technology (Rogers, 1983:165). Theoretically, it is at the fifth stage of the innovation- decision process that the impacts of the technologies will be felt and will feed back into the model either to confirm or eventually lead to the rejection of the technologies. Rogers claims that discontinuance is more frequent when innovations are less compatible with the individual's beliefs and past experiences (Rogers, 1983:188). He states that ”90 percent of all new products fail within four years of their release" (Rogers, 1983:211). The diffusion model recognizes the interaction of values and technology with the assumption that potential adopters' values are harnessed by change agents to increase the adoption potential of a new technology. The Lovinses have (performed the function of change agent through efforts to show how the use of renewable energy technologies will more 32 fully implement the values held by the American population. Their emphasis on the social impacts produced by the renewable energy technologies have differentiated them from most diffusion change agents. Very few diffusion researchers have examined the consequences of innovations. This is true because traditionally positive results of the adoption of an innovation are assumed and because "consequences are difficult to measure" (Rogers, 1983:378). Rogers, however, does show sensitivity to some of the impacts of innovations on social stratification systems. A system's social structure partly determines the equality versus inequality of an innovation's consequences. When a system's structure is already very unequal, it is likely that when an innovation is introduced (especially if it is a relatively high-cost innovation), the consequences will lead to even greater inequality in the form of wider socioeconomic gaps. (Rogers, 1983:402) He hints at the importance of concomitant restructuring of society if technology is to have appropriate social impacts as he comments on diffusion research: A means to social revolution it is not. A helpful tool for social change and development, when accompanied by a basic restructuring of society, it may be. (Rogers, 1983:125). 33 He therefore questions the feasibility of using technologies to create changes which are different from the dominant patterns of society, such as using solar energy technologies as a tool for changing to the soft energy path. Considerable work has been done out of the diffusion tradition on energy technologies and practices and on solar energy technologies. The Diffusion of Energy Conservation Practices: The Lovinses have stressed steps in the change to the soft energy path. The first step is to use energy more efficiently (e.g. energy conservation) and the second is to use renewable energy technologies to provide a new energy base for society. As shown by the early theorists of energy, technology and social change, a close association was assumed between the amount of energy consumed by society and the degree of industrial development of that society. In fact the amount of energy per capita use was considered a primary indicator of economic development (Hirst et al., 1983). As empirical societal comparative studies were conducted, it became clear that energy use and industrial development are not as tightly coupled as previously thought. Studies of Sweden and Germany (Schipper, 1982; Lonnroth et al., 1980) were instrumental in severing the assumption of tight coupling. 34 These studies indicated that there were European countries where the level of industrial development and quality of life were as high or higher than that found in the U.S.; yet the amount of energy consumed per citizen was much smaller. From 1960 to 1973 both GNP and energy consumption growth increased at proportional rates in the U.S. However that link was severed after the 1973 oil embargo and from 1973 to 1980 the energy use rate increased at a lower level than the GNP rate. By 1983 less energy was used per GNP dollar than in 1974. Conservation was responsible for about 5 percent of the energy savings in 1980 (Axelrod, 1984:214: Hirst et al., 1983). Several researchers examined how energy conservation occured in households. In a study conducted in 1978, Kempton et a1 (1982) examined the perceptions of energy conservation among Michigan householders. Responses were placed in three categories of energy conservation: (1) efficiency investments, (2) management (e.g. turning down the thermostat) and (3) sacrifice of amenities or comfort. Using open ended telephone interviews, the researchers found that the three Inost commonly used measures were, "lighting reductions, leewer thermostat setting, and adding insulation" (1982:6). WRJey also found that the respondents overestimated the savings of "sacrifice and management" while underestimating 35 the savings of "efficiency investments" (1982:11). In further examining the actual use of efficient energy technologies, Schipper (1982:10) found that consumers "seem uninterested in anything that takes more than five years to pay back". Williams et al. (1983:284) set the range at 2-4 years. Furthermore, Goldstein (1983) found that customers would not invest an extra $150 to buy a more efficient refrigerator model which would give them a $600 savings over the life of the appliance. The first step advocated by Lovins for changing to the soft energy path has been substantially implemented. Energy conservation measures have diffused quite successfully through the American society (Hirst et al., 1983; Williams et al., 1983: Farhar-Pilgrim and Unseld, 1982: Levine and Craig, 1985). In fact Schnaiberg (1983) has contendgd that one of the few successes of the soft path advocates has been to increase conservation. In a study of public acceptance of energy conservation strategies, conducted in Washington State, 1981, Olsen (1983) found that soft path preferences, measured by Preferences for policies based on renewable energy resources and conservation as opposed to oil, coal and nuclear Sources, was the strongest predictor variable for the acceptance of energy conservation strategies. The second 36 strongest predictor was perceived seriousness of the national energy problem as measured by the questions "Do you consider meeting the United States' energy needs during the next ten to twenty years to be: not a serious problem, a somewhat serious problem, a serious problem, or a very serious problem?" (Olsen, 1983:190). Olsen reports that his Soft Path Preference Index was most strongly related to energy conservation strategies which involved voluntary community programs of conservation, land-use changes through zoning to discourage urban sprawl, and setting consumption limits based on energy supplies. Those perceiving energy as a serious problem were more likely to see efficiency standards as the primary method of increasing energy conservation. Lovins (1976, 1977) claimed that conservation is a measure economically important for all to take no matter what the policy preference. Both the soft and hard path advocates agree that energy conservation is important. She Diffusion of Solar Technologies: The second and most important step in a change to the to 80ft energy path is the implementation of flow energy base Camanging to renewable sources of solar, wind, and water. Public opinion surveys and reviews of the early 1980's 37 indicated the same result: the public favored solar energy technology developments more than nuclear technology and nonrenewable energy source developments as the energy base of the future (Yankelovich and Lefkowitz, 1980: Farhar et al., 1980: Farhar-Pilgrim et al., 1979: Mitchell, 1980, Mitchell, 1984, and Olsen et al., 1982). In fact Olsen et al. summarize the data by saying: The conclusion to be drawn from these recent studies is unequivocal. A substantial majority of the American public opposes further development of nuclear power plants. Conversely, a large majority of the public supports efforts to promote greater energy conservation, especially in the residential, commercial and industrial sectors...And almost everyone favors increased use of solar power. (Olsen et al., l982:5) Although the public favored the renewable energy base for a future time period, it is not clear that it is moving to the implementation of such a change in the present. Several researchers have suggested reasons for the lag between attitudes and behaviors. Yankelovich and Lefkowitz (1980) reported that the level of trust that the American public has in technology's ability to find solutions to problems of shortages and natural resources was down to 52% from nearly universal support for technological solutions after WWII. The confidence in technology ranged from 69% confidence among the older, less Well-educated lower-income segments of the population to a 38 29% confidence among college students. Bezdek et al. (1982) suggested that one of the barriers to making solar energy 20% of the national energy base by the year 2000 was that the public has "lack of confidence in solar energy technologies” (Bezdek et al.,l982:339). The public opinion surveys previously mentioned indicate questioning the feasibility of implementing solar technologies. While supporting the increased personal control of the availability and type of energy use and the decentralized energy production, Bezdek et al. propose achieving those goals ”will require a major national commitment to resolve the obstacles hindering the development of solar technologies" e.g. reliable solar equipment for some technologies, high capital cost, existing legal and institutional barriers, lack of market infrastructure and manufacturing capabilities, and lack of public knowledge and confidence in the technologies (Bezdek et al., 1982:339) Bezdek et al. used several models to examine the implications of three levels of solar technology penetration into the U.S. markets. They concluded that because solar ‘technologies "require higher initial capital investment per talit of energy produced than do alternative, conventional energy sources” the amount of private and federal 39 expenditures required to put them in place would range from $450 billion to $1.2 trillion (Bezdek et al., 1982:349). They argue that this support was not feasible given that a shortage of capital was perceived as one of the most serious problems facing the nation in the 1980's and 1990's. The Mitchell (1980) data indicated that 28% of the public nationally had considered installing a solar system while 1% had actually done so. About 0.5% of the American population reported owning solar heating systems in 1979, showing no change since 1978 (Farhar et al., 1980:160). In examining the market penetration of solar systems, Roessner concluded that 1.5% was the maximum penetration of solar technologies in California, this figure probably representing one of the highest penetrations in the nation. Nationally, penetration was very low, "in the order of 0.3% of the maximum potential U.S. market by 1980. In the national residential market, penetration can be said to have barely begun" (Roessner, 1982:10). The diffusion rate is less clear in studies of passive solar technologies. Nelson and Honnold (1980) examined the diffusion of passive solar designs in an attempt to discover the pattern of a technology which is not centralized in design or promoted by a social change agent. They suggest 40 that perhaps people who adopt decentralized systems are different from those who use traditional energy systems or systems with active solar components. Nelson and Honnold concluded that "bureaucratic action has retarded 'passive technologies' diffusion in the past" (Nelson and Honnold, l980:5). However the meaning of "bureaucratic action" was not specified further. Other researchers have suggested that the nature of passive solar technologies gives little incentives for corporate or governmental involvement (Davis, 1982: Williams et al., 1983). The diffusion of passive solar technologies was examined by Eastman (1982) in a study of "a breadbox solar water preheater, a window box heater, and a retrofit trombe wall" (Eastman, l982:l) which were introduced to low income residents or public facilities buildings in New Mexico. A follow-up study of the 381 workshop participants was conducted two years later. Forty of the 122 who were interviewed had installed some kind of solar device since the workshop - greenhouses, trombe walls, solar water ‘heaters, and some type of solar home. All the adopters expressed a high degree of satisfaction with their devices, though few could estimate their energy savings (Eastman, 1982:4). 41 However, with ”two exceptions and despite effusive praise, the solar demonstration recipients were either not using or were benefiting very little from their solar devices" (Eastman, l982:8). Eastman concluded that perhaps one of the reasons why the technology failed to diffuse was that the recipients were, with two exceptions, aged, infirmed or socially inactive. Therefore, although public opinion was very strongly in favor of changing the energy base of society through solar energy technologies, it is not clear that the households were adopting one type of renewable energy technology necessary to change the energy base. CHAPTER III RESEARCH RELEVANT TO DIMENSIONS OF SOFT PATH PREFERENCES In this chapter I discuss the social context in which Lovins' ideas became influential. Then I review research specifically relevant to the dimensions of soft path preferences. The dimensions are the central values which form the core of a consistent package of attitudes and behaviors which are congruent with soft energy path structural changes. The dimensions are: (1) natural resource conservation preferences, (2) self-reliance preferences (e.g. participation in decision-making), (3) personal gains preferences, (4) cross-generational, crossnational, and cross-class equity preferences, and (5) social diversity preferences. THE SOCIAL CONTEXT OF THE LOVINS PROPOSAL Lovins' seminal paper (1976) comparing the hard and soft energy paths was published at a very propitious moment in history. Morrison (1980) traces the enthusiasm of environmentalism through the 1960's, peaking on Earth Day 1970. The 1973 oil embargo highlighted national and iJTternational equity issues of resource constraints, sensitizing some environmentalists to the potential equity 42 43 impacts of their concerns. The appropriate technology movements's founder had already published his catalytic book (Schumacher, 1973) identifying technology as a force of inequitable international development. Schumacher's assertion of two broad technology types driving different types of social changes was permeating the thinking of scholars and social activists. The inappropriate technologies had characteristics compatible with the hard path technologies criticized by Lovins. The appropriate technologies were more broadly focused (e.g. crop production, water, etc.) than the soft energy technologies proposed by Lovins but had similar characteristics . The anti-nuclear movement was also gaining momentum. It supported soft energy technologies as a positive alternative to nuclear power (Morrison, 1980)7. These historical events plus the changing cultural values of Americans produced a very responsive social environment for Lovins' proposal. Changing Cultural Values: Yankelovich and Lefkowitz (1980) record significant changes in Americans' perceptions of economic growth by noting: 44 From the 1950s to the late 19603 Americans characteristically believed the present to be a better time for the country than the recent past and anticipated that the future would inevitably improve over the present. In 1971 the pattern changed. Then Americans saw the past in a rosier light than the present, but anticipated that the future would once again brighten up for the country. In 1978 for the first time the pattern of the 50s totally reversed itself. Now Americans believe that the past was a better time than the present, and they anticipate that the present, however had, is likely to be better than the future. This, indeed, is a historic shift away from traditional American optimism to an uncharacteristically bleak outlook. (Yankelovich and Lefkowitz, 1980:102). The historical cultural assumption that the country could provide material stability and security for its populace had been challenged by: (1) a redefinition of success to include more nonmaterial standards of personal growth: (2) a pressing for safer and healthier environmentS; (3) a pervasive distrust of the dominant institutions of society, especially big business and government, and (4) a questioning of U.S. continual dominance in the world economy. There had been a shift in expectations. However, people still hadn't come to terms with the "conflict and disappointment created by the need to adapt to new, unwelcome conditions" (Yankelovich and Lefkowitz, 1980:99). Value changes interacted with the "energy 45 crisis" to create a very unstable situation. Others writing around the early 1980's also were concerned with the instability of the cultural system. Anderson (1983) and others (Schurmann, 1983: Morrison, 1983, 1980) suggest that new choices were being made cutting across liberal and conservative frames of references. What appears to be happening in the world is a . simultaneous march of events in opposite directions: Going one way, a parade of decentralist and separatist movements - there is scarcely a national government in the world that is not struggling with one or more of these - and, going the other way, an unprecedent increase of global trade, global communication, global migration, global interdependence. Things flying apart and at the same time coming together. (Anderson, l983:6) Carlson et al. (1982) conclude that "the societal incentive system tends to be out of synchronism with shared social interests and individual concerns" (Carlson et. a1, 1982:155). Lovins' presented his proposal in the midst of society's turbulent inconsistencies. The ”Energy Crisis": The "energy crisis" reflected and fed the unstable characteristics of the period. Even while people were slowly believing that the "energy crisis" was real - as opposed to something contrived by "big business and government" - there was fear about facing a reality which would be increasingly bleak. As Yankelovich and Lefkowitz (1980:110) express it: 46 "In short, the public is caught between two feelings: the emotion that 'somehow it cannot be' and the emotion, 'my God, it may be worse than we think'." Energy is one of the first forces of the modern era to create a confrontation with limits. Schnaiberg (1983) argued that this made energy technologies more visible. The ambivalence of the population's reactions was documented in the research findings of the period. Farhar et al. found that "most people do not believe there is an energy crisis but perceive instead a serious national energy problem". created by big business (especially oil companies) and government (Farhar et al., 1980:143). In a national study, Farhar-Pilgrim and Unseld (1982) indicate that homeowners did not perceive serious impacts on their households' lifestyle. On the other hand, Olsen et al. (1982) report that perceptions of the U.S. "energy situation as very serious" were associated with support of energy conservation as a national energy policy (Olsen et al., 1982:7). Also, concerns about energy shortages were second only to inflation in an opinion survey commissioned by the Council on Environmental Quality (Mitchell, 1980). Although there were diverse findings, public opinion 47 studies of the 1970's and early 1980's indicate a general prevalence of perceptions of system vulnerability generated by energy issues (Schnaiberg, 1983;Levine and Craig, 1985). A sense of urgency to resolve the uncertainties generated by perceived system vulnerability was captured by Amory Lovins' presentation of the dilemma as a choice between a "hard path", representing the 91d vulnerable order, and the "soft path" which represented a positive response to the vulnerability. Lovins (1976, 1977) urged a soft path social policy choice soon, a choice that would be implemented by individuals and local social organizations' decisions to respond to economic and noneconomic incentives to change. The social structural changes would occur as the choices aggregated. Carlson et al., (1982) urged similar decision-making speed about the societal energy strategy. The longer it takes society to make up its mind with regard to its ultimate dominant energy strategy,the more difficult will be the implementation of whichever path is chosen. (Carlson et al., 1982:151) In later writings (Lovins and Lovins, 1982; Lovins et al., 1983), local communities were especially urged to become the key decision-making units to choose the soft path. The Solar Technology Choice: Given the prevalent perception of system vulnerability, were solar energies being chosen as tools for resolving that 48 dilemma? The recorded diffusion of solar energy technologies was, as noted previously, slow. Even if solar technologies were chosen as a strategy for changing the societal energy base, some reseachers doubt that the technologies will contribute to further cultural value changes (Barbour et al., 1982: Hornick and Enk, 1980). Yet others propose value changes as a likely social impact of solar energy technologies (Piernot et al., 1981). The interaction of solar energy technology use and cultural values is not clear. This project is a contribution to examining this linkage. RESEARCH ON THE DIMENSIONS OF SOFT PATH PREFERENCES While not directly related to the theory of soft path social change, some relevant studies about the claims of the soft path preferences are reviewed to assess whether or not they support Lovins' claims. Natural Resources Conservation Preferences: Several studies (Mitchell, 1980; Farhar et al., 1980:153) have reported that opinions about energy-environment trade- offs were polarized, with sizable minorities favoring each side. At the same time, there was an inclination towards valuing an adequate energy supply. 49 Mitchell (1980, 1984) traces the emergence of environmental concerns through the 1970's and the stability of those concerns into the 1980's. He claims that in the 1970's, the environment was perceived as being in crisis. By the early 1980's, the polls indicate that the American public no longer regarded it as being in crisis but was still supportive of protecting the environmental (Mitchell, 1984:10). In the 1980's a clean, healthy environment is perceived as citizens' right or "entitlement" so therefore it is less salient, he argued. In 1982, people perceived that the environment in the U.S. "had grown worse" (48 percent) compared to 10 years previously. Only 34% felt that it had improved (Mitchell, 1984:14). Thus the risks of environmental decay are a public concern. The reliability and risks of solar technologies have been addressed in several different ways. Holdren et a1. (1980, 1982) have compared the environmental risks imposed by renewable and traditional energy technologies. They conclude that ”the use of passive solar design in architecture will produce smaller environmental impacts than those resulting from supplying an equivalent amount of energy from any of the 'active' technologies" (Holdren et al., 1980:249). 50 The land use requirements of solar heating and cooling systems have been difficult to anticipate. However, if neighborhood-scale solar energy systems and district cogeneration were used, then a high demand for land would not occur. Furthermore, land used for decentralized solar energy technologies could also be used for other purposes. Bezdek et al. (1982) consider the demands for land as one of the larger drawbacks to the development of solar technologies. While admiting that in the long run solar technologies are less polluting and resources using than other energy sources, Bezdek et al. (1982) note that initially they are resource-intensive and pollution creating thus degrading the environment in the short run. The amount of degradation posed by different types of renewable energy technologies was more closely examined by Holdren et al. (1980). They indicate that the potential for catastrophic accidents among the renewable technologies could only occur with large, centralized facilities, e.g. large hydro-electric dams and satellite power stations. Irrigated biomass plantations are the renewable energy forms with greatest potential for producing major climatic impacts. Although these three technologies use the renewable energy base, they do not have the other properties of "soft" energy technologies especially, they are large scale, 51 centralized, and not in control of the end-users (Lovins et al., 1983). Lovins and Lovins (1981) argue that the way to reduce current worldwide air pollution is to use solar energy technologies even without changes in lifestyle and assuming full industrialization of countries which are not currently industrialized. Noncatastrophic characteristics become more important as issues of the catastrophic potentials of highly complex interdependent systems are more salient (Perrow, 1984). The relationship between the amount of environmental degradation per unit of energy produced has been the focus of researchers from the Department of Energy (Bezdek et al., 1982) and from the academic sector (Holdren et al., 1980) who reached similar conclusions. Some renewable energy technologies have the possibility of reducing environmental costs per unit of energy produced to well below those which have been produced by oil and coal. The more benign renewable options are: (l) passive solar heating and cooling, (2) increased electricity generation by adding generators to some of the existing dams, (3) electricity generation by wind turbines, and (4) biogas digestion of sewage and feedlot manures (Holdren et al., 1980:283). The health implications of developing renewable energy 52 technologies have also been examined. Inhaber (1979) proposed that the health risks of renewable technologies are "much greater than those of natural gas and nuclear power and comparable in many cases to those of coal and oil" (cited in Holdren et al., 1980:267). Holdren et a1. reach a much different conclusion concerning a comparison of such systems, namely the total occupational effects for all stages of the renewable energy system fuel cycles conceivably could be equal to the total occupational effects of obtaining an equivalent amount of energy from coal. (Holdren et al., 1980:270) The technical and economic risks and reliabilities of mixing wind and solar systems with conventional energy systems have also been examined. Kahn concluded that "exogenous, unplanned risks have a smaller impact on the wind energy system than on the conventional one" (1979:343). Further, unrestricted entry into decentralized application of wind and solar energy conversion will have a destabilizing effect on the electric utility industry (Kahn, 1979:347). "Wind and solar energy are more economically competitive at lower standards of reliability" (Kahn, 1979:337) than conventional energy systems. But even in California, with a 55% state tax credit for solar systems, this only partially "offsets the subsidies to conventional heating sources" (Kahn, 1979:345). Thus Kahn agrees with Lovins' claim that solar technologies are more economically efficient than 53 traditional energy technologies, but face more restrictive barriers. As seen above, these findings generally support Lovins' claims that renewable energy technologies are more environmentally benign and resilient than nonrenewable energy technologies. They also suggest the potential for conflict with the utility companies (e.g. controllers of the hard energy system) of a change in societal energy base. Self-Reliance, Decentralization and Grassroots Democracy: Another dimension central to the development of the soft energy path is the belief in the value of public participation in governance. . .or "grassroots democracy" rather than reliance on centralized governmental institutions. Yankelovich and Lefkowitz (1980) record that a dramatic change which occurred during the decades of the 1960's and 1970's was a decline of trust in government. This was reflected in the findings that the government and oil companies were perceived as most responsible for the nation's energy problems (Farhar et al., 1980; Farhar- Pilgrim et al., 1979; McFarland, 1984; McKie, 1984). Mitchell (1980) also records the predominance of public sentiment that government does not provide ways for expression of their thoughts about the environment, and that 54 interest groups, which are perceived as being separate from the general population, had greater influence with the government. Concerns with personal self-reliance were shown in Yankelovich and Lefkowitz's (1980) findings that people were increasingly seeking to claim control over their own destinies. McFarland (1984:514) suggests that the U.S. is "probably the industrial democracy that most highly values local political participation and policymaking, as opposed to the politics and policies of the central government." These public opinion polls and reviews support Lovins' strategy of noncoercive participative local control of energy systems (Lovins, 1977; Lovins et al., 1983; Morrison and Lodwick, 1981). The meaning of decentralization and self-reliance is examined with the question, "how much power to which people?” Cose suggests the following self-reliance options: (1) let the marketplace function without barriers by reducing the roles of local and federal governments; (2) give federal resources and responsibilities to local governments; (3) turn the responsibilities to solve problems to corporations or other private entities (Cose, 1983:11) . fluie Lovinses propose following strategies one and two (e.g. allow the market to function freely and local communities to 55 have the resources and control over their own energy developments). Decentralization of the energy supply and the resulting self-reliance is viewed by the Lovinses, as a positive result of the implementation of the soft energy path in the U.S. The federal government was providing some incentives for such developments in 1980. The impacts of these incentives on increasing energy conservation and the diffusion of solar technologies have been examined by several researchers. One of the early studies of the effects of federal incentives was conducted by Seymour Warkov to examine whether or not the HUD Solar Hot Water Grant Program was able to speed the adoption of that solar heating and cooling technologies with a $400 grant. The HUD Program was started in 1975, and by 1980 HUD had funded the installation of solar space and or domestic water heating systems in nearly 12,000 houses or apartments buildings with a total expenditure of more than $21 million (Roessner, 1982:6). Warkov found that the most important variables for adoption of the technologies were support from personal networks and perceived private benefits for the Connecticut homeowners. This is fully compatible with the Lovins' notion. Warkov also found that "measures of knowledge about legal and 56 financial incentives, etc., and evaluations of HUD program effectiveness did not predict household adoption of this technology" (Warkov, 1979:ii). Roessner (1982) further reviewed the results of governmental strategies for solar commercialization in effect from 1974 to 1980. The National Energy Plan was the foundation for governmental strategies of making the energy economy a "joint public/private concern and responsibility" (McKie, 1984:344). It was incorporated in the five bills of the National Energy Act (NEA) passed by Congress in 1978 and later. The NEA provided income tax credit of 30% of the cost of solar energy technologies up to $2,200. This was increased to 40% credit for a maximum expenditure of $10,000. About 35,000 people claimed solar energy tax credits in 1978, costing the government $32 million. This figure rose to $44 million in the 1979 tax year (Roessner, 1982:6). Beginning in 1975 and 1976, several states enacted legislation creating solar financial incentives. The largest was in California with a 55% tax credit (Roessner, 1982: Levine and Craig, 1985). Others are listed in Farhar- Pilgrim and Unseld (1982). These incentives were based on the assumption that economic factors were the primary barriers to the dissemination of solar energy systems (Roessner, 1982). 57 In a study of ten Western states Carpenter and Chester (1982) found that 63% of those who had installed solar water or space heating technologies claimed that they would have done so without the tax credits. The authors concluded, that "for large expenditure energy conserving devices a tax credit appears to be required" (Carpenter and Chester, 1982:10). This finding was also supported by Farhar-Pilgrim and Unseld (1982). These adopters, however, are early adopters who generally have many socioeconomic resources. The studies reviewed above focused primarily on active solar energy technologies since they were the ones supported with governmental incentives in 1980. Vine differentiates between active and passive solar adopters in his study of members of solar energy voluntary citizen groups. These people had more passive solar systems than active or mixed systems (Vine, 1980:132). He suggested that their prime motivation in acquiring the technologies was economic while saving energy was the most common second reason. "Very few [less than 10%] installed solar because of their support for the idea of solar energy or because of their desire to become less dependent on the utility company" (Vine, 1980:132). In New Mexico, installers believed governmental rebates encouraged solar use; builders (who were selling passive solar designs) did not find them useful (Eastman, 1982). 58 Generally, then, governmental policies in existence during the ”energy crisis decade" were not very effective in speeding early household adoption of solar energy technologies. The effects of governmental incentives on energy conservation were also investigated. Shippee (1980) found that information, attitudes toward the energy crisis and demographic factors were very low in predicting energy conservation. Financial incentives, depending on their timing and amounts, were important, as was feedback tied to committed efforts to reduce energy consumption (Shippee, 1980). On the other hand, Carpenter and Chester (1982) found that 86.8% of their respondents in Western states were aware of federal tax credits for energy conservation practices. Of these, 34.5% had made a claim but only 6% said that they would not have taken the conservation measures without the tax credits. This, generally, was the result of many studies. Conservation measures, especially the less costly ones, have been implemented in residential homes without tax incentives (Williams et al., 1983; Levine and Craig, 1985:575). Roessner concluded his analysis of the effects of federal 59 and state tax incentives by saying: Studies of the consequences of state solar financial incentives and of the factors that influence homeowners to purchase or not purchase residential solar heating systems provide evidence that environmental concerns, self-reliance, and propensity to innovate are at least as important, and may be more important, than cost savings in decisions to diffuse solar heating systems during the early stages of market penetration. These findings indicate support for Lovins proposed importance of noneconomic incentives for the acquisition of solar energy technologies. (Roessner, 1982:18). These findings indicate support for Lovins proposed importance of noneconomic incentives for the acquisition of solar energy technologies. Some communities have passed local codes to encourage energy conservation and the use of alternative energy sources. Davis, California is a classic case. Dietz and Vine (1982) assessed the impacts of an Energy Conservation Building Code to determine if the code reduced the amount of energy used. Although there had been a significant reduction in energy consumption, the researchers could not determine the influence of the code because it was too small during the 1970-1979 period. Besides examining issues about the role of the federal and state governments in creating economic and legal incentives for decentralized energy technology adoption and 60 conservation, researchers have analyzed the effect of feedback to the government from the population. Alford and Friedland (1975) warn about the dangers in assuming that power and participation are associated. ”Participation may be associated with power, but power can exist without participation" (Alford and Friedland, 1975:430) through the creation and maintenance of social structures which reinforce the power of elites. It is the elites' "power without participation" which the Lovinses want to change through "power with participation" of the populace. Again Alford and Friedland warn that "for nondominant interests, participation has been a symbolic substitute for power, a means of reproducing the absence of political power" (Alford and Friedland, 1975:474). McFarland empirically examined the role of energy lobbies such as the Solar Lobby in influencing federal policies. He concluded that most of the governmental policies of solar power had been based on an "enthusiastic President with a positively disposed Congress" during the Carter administration and an antagonistic President and advisers during Reagan's administration. During neither period were the changes based on power vested in citizens' interest groups. Congress perceived solar energy and energy conservation as being popular with the voters (McFarland, 61 1984:519). Lovins and Lovins (1982) have argued that energy decisions should be made at the community level. They claim that there is a sense that when there is another disruption in energy supplies "it will be every community for itself" (Lovins and Lovins, 1982:301). Community-based action “is the fastest and surest way to build a resilient energy system" (Lovins and Lovins, 1982:332). Stern and Aronson (1984) suggest that local solutions may simply be easier to achieve than national solutions. Their commitment to local communities was the source of the Lovinses' criticisms of the Carter administrations' approach to the energy crisis. Carter used a "permanent emergency" authoritarian institution to implement changes rather than noncoercive grassroots participation (Lovins et al., 1983:66). Communities ”are the level at which, at least in energy policy, solutions are possible." Community is in the unique position of being large enough to mobilize resources needed to meet such a threat as energy but small enough to understand and protect the community's diversity, to allow for constructive individual participation, and to tailor the solution to meet individual needs. (Lovins et al., 1983:75) The federal government should help communities serve the national interest by implementing their own plans for energy self-reliance propose the Lovinses (Lovins et al., 1983). 62 This preference for community based energy decision-making has been strongly criticized by some social scientists. Local developments can lead to conflicts between neighboring facilities and will probably not meet the needs of society as a whole, they warn. Programs implemented under a set of local conditions cannot be transmitted to another set of conditions (Stern and Aronson, 1984:166). Other criticisms have focused on the Lovinses' assumption that small technologies will necessarily lead to greater democracy. Self-reliant and decentralized technologies may need to be managed in a very centralized manner (Gerlach, 1981:57). They may require mass production and distribution systems or centralized control (Barbour et al., 1982). Wood stoves, for example, are decentralized technologies with central control, as illustrated by Denver, Colorado's passing legislation to prevent their use on high pollution days (Kowalski and McBean, 1986). The national public good may be threatened by community- based energy decisions/policies. Turning over all energy policy formation to state and local governments is "unrealistic" for they cannot deal competently with things such as oil import quotas. In this Cose (1983) agrees with Alford and Friedland (1975) who claim that 63 The decentralization of funds, policy, and implementation discretion to the city level of government thus reinforces the urban polity as the unit of political participation for response to problems whose causes lie outside the urban system; this insulates from any kind of political challenge dominant interests whose political power and economic organization are located elsewhere. (Alford and Friedland, 1975:464) Personal Gains Preferences: Lovins (1977) clearly advocates free market transactions for distributing renewable energy technologies. The ability of the marketplace to distribute solar energy technologies has been examined primarily through studies of the role of "big business" in the solar energy market. This focus is consistent with Barbour et al.'s (1982) proposal that one way individuals can participate in decision affecting their lives is to have the freedom from the pervasive power of large organization's control over productive activity. Reece (1979) was one of the first to systematically examine the evolving centralization of solar energy markets. He and others (Levine and Craig, 1985; Ridgeway, 1982) report that the early emphasis of the federal government was on large scale solar energy technologies such as satellite solar technologies and a large scale power tower in California's Mojave Desert (Ridgeway, 1982) rather than on home heating (Reece, 1979:76). Furthermore, Reece documents how large corporations have "concentrated sufficient control over the 64 solar industry to squeeze out smaller competitors and effectively prevent the entry of others" (Reece, 1979:187). Dietz and Hawley (1982) have explored the case of photovoltaics, concluding that ". . .the future of U.S. [photovoltaic] industry innovation is in the hands of large diversified, and primarily oil, corporations. . .since the late 1970's through mid-1982 there has been an apparently high concentration ratio among U.S. producers" (Dietz and Hawley, 1982:25). They indicate, that the photovoltaic industry is a global industry (Dietz and Hawly, 1982:25) with primary markets in Third World countries and in competition with other countries, particularly Japan and European nations. Most of the photovoltaics were subsidized by parent companies, thus seeming less costly than they actually were. Global corporate members feared that the companies would go out of business unless there was an improvement in the prices at which they could sell their products. .Purdy (1985) also documents the increased concentration of ‘traditional energy corporations and of their control over <5ther energy resources including 75% of the solar industry (Purdy, l985:9). For example 65 petroleum corporations own five of the nine leading solar energy firms, with Exxon and Atlantic Richfield owning half of the solar photovoltaic cell industry. (Purdy, 1985:16) Through their diversification and wealth translated into power, the individuals at the pinnacle of these corporations "make decisions concerning the size and rate of growth of many sectors of the economy, as well as the direction and implementation of various types of technology" including, especially, active solar systems and photovoltaics (Purdy, 1985:28). Ridgeway (1982) further examined the centralized controls over solar energy technologies concluding: Even Government programs for demonstrating solar collectors at homes throughout the country have been steered to the larger firms. Almost 70 per cent of the $6 million available through the Department of Housing and Urban Development's solar demonstration program for 1977 went to seven major corporate solar manufacturers. All but two were solar subsidiaries of large American corporations, including Exxon, Aarco, and Grumman. Some of the long-time solar pioneers were excluded from the program because, ironically, their systems were too cheap (Ridgeway, 1982:346). The solar energy technologies have not only been coopted by large corporations, but the political control of energy has also become more centralized. Davis (1982) examines the .interaction of the development and control of energy industries based on their political context and on the 3physical characteristics of different types of energy 66 sources 0 He suggests that coal is one of the least regulated industries because of the historical period when it developed, while natural gas is the most regulated one. "The political process determines the issues of ownership, prices, and quantity consumed, which in coal and oil are decided privately" (Davis, 1982:164). Davis claims that the "ecological ethos" of the current era has created a situation where the "new fuels evoked quasi-religious support. Sunlight, geysers, and the wind were natural, God given and nonpolluting" (Davis, 1982:244). These fuels are symbols of environmentally benign living. The difference between the centralized control of active solar technologies and the decentralized local control of passive technologies is analyzed by Davis. Since solar power, especially passive, can be provided by a single homeowner, there are no natural monopolies "requiring government supervision nor. . .any advantage for large corporations" (Davis, 1982:246). Thus passive solar technologies' small scale and flow characteristics encourage no federal governmental involvement. 67 Yet, since 1975, there has been bureaucratic and legislative reorganization centralizing energy politics and shifting energy problems into the political system rather than into the economic system (Davis, 1982:283). McKie (1984) cited market failure as a reason why the federal government has attempted to regulate the energy economy. Issues of national security, equity, threats to the environment, and long term needs are not adequately addressed by market mechanisms (Williams et al., 1983: Craig and Levine, 1985). However in 1981, there was a partial return to the market with Reagan's energy policy. Reagan was unlikely "to entertain any more claims of 'market failure' to justify new regulations" (McKie, 1984:346). Ronald Reagan's free-market approach to energy policy has been condemned for reducing financial support for renewable energy sources while continuing support for nuclear energy (Green et al., 1984). The 1984 budget requests for solar and renewable energy showed that the hardest hit programs were the solar programs being reduced from $61.1 million to $21.9 million (Axelrod, 1984:211; Green et al., 1984: Levine and Craig, 1985). Reagan's argument is that solar and renewables programs have been so successful that they no longer need governmental support (Axelrod, 1984:212). This review has not found documentation for the success of the diffusion of solar renewable technologies. 68 Equityqlmpacts: One of claims made about the social impacts of using soft energy technologies is that there will be more natural resources available for everyone than if hard energy technologies are used. Therefore, there will be greater equity within and across nations and generations (Lovins, 1977: Lovins, 1980; Lovins and Lovins, 1982; Lovins et al., 1983). When energy resources are scarce or expensive, people with few economic resources are more adversely affected. Unless there is a change in the energy base of society, some may have to choose between eating and keeping warm, claim the Lovinses (Lovins et al., 1983). Cooper et al. (1983) document how low income groups' energy expenditures rose from 11% of their income in 1972 to 23.2% percent in 1981 while the higher income households' expenditures increased from 2.5% of their income to 3.5% during this same period. A review of the energy literature records how specific low income groups experienced the "energy crisis". Older people reported more adverse effects of the energy situation than younger people; nonwhites experienced more negative financial impacts due to energy shortages than ‘whites (Farhar et al., 1980). Cooper et al. (1983) summarize the situation by suggesting ‘that lower income houses have already cut back on energy use. If future energy price increases occur, there will 69 be greater vulnerability in that segment of the population than what it experienced in the past. Although many of the Lovinses equity concerns are expressed in relation to cross-national and cross-generational issues, the impacts on lower income classes are consistent with their equity concerns. Critics have charged that decentralizing energy policies will decrease equitable access to energy resources (Cose, 1983; Stern and Aronson, 1984: Barbour et al., 1982). Social Diversitinmpacts: While social diversity impacts have not been directly addressed, they have been assumed by critics of the soft path social change theory. Most of the criticisms have centered on the potential transitional or permanent conflicts engendered by a change to the soft energy path. Suggestions that decisions made at the household level could have negative effects on centralized energy institutions create a potential for conflict. Gilmer (1980) examined the historical, technical and economic advantages of central station power generation. He concluded that should those competitive advantages change "either because of :xncertainties plaguing the utilities or advances in decentralized solar technology, these organizational 70 advantages and the simplicity of social control may begin to weigh heavily in favor of solar energy" (Gilmer, 1980:23). The utilities would be the losers if many consumers began relying on their own power systems. Energy base conflicts have been more closely examined by several researchers. Morris (1982) claims that the transition period will not be smooth or painless as implied by Lovins. "Energy wars" have already broken out. Communities have reacted against attempts of utility companies to increase distribution of electricity to urban centers. Gerlach (1981:1982) examined some of the "energy wars" and found that the parties involved in the disagreements used whatever rhetoric was most supportive of their political purposes. For example, the soft path rhetoric was used by farmers to support their position against large energy companies but was changed when doing so was more politically advantageous. The theme of the conflictual reorganization of society, has been strongly argued by Schnaiberg (1983), Perelman (1980), and Perelman et al. (1981). Schnaiberg suggests that it is not correct to argue that opponents will simply accept the necessity for changing their values and behaviors. The change to the solar path is a change in the production 71 system of society. Such a change provides an enduring base for conflict as it touches on "the economic interests and political influence of powerful economic interest groups" (Schnaiberg, 1983:229). The Lovins' theory must be broadened to include social conflicts and social interests. As part of this, Schnaiberg suggests that it is important to distinguish between ". . . changes in the forces of production (physical technology) from those in the relations of production (social class structure)" (Schnaiberg, 1983:229). Scholars (Stern and Aronson, 1984: Cose, 1983: Perelman, 1980) have concluded that placing energy decision-making in the hands of diverse decentralized areas would increase the conflict between neighborhoods, communities, states and nations of the world, contradicting Lovins. Cose (1983) speaks very forcefully of conflicts engendered at the community level. However, citizen groups, despite loud and visible protests, are incapable of threatening communities as much as major corporations which propose witholding investments or leaving. He concludes 72 The specter of conflict . . . looms over the land . . . perhaps more so than at any time since the Civil War. Energy is not the only cause of conflict between states. Competition over defense dollars and corporate investments accounts for much of it. But a number of disputes have centered around energy. (Cose, 1983:86) Perelman agrees: The transition which will occur in the course of the next century or so is likely to be a period laden with intense social conflict, and probably violence. Exactly what kind of society will emerge from this transition period is an extremely speculative question, subject to all sorts of alternative possibilities in light of unstable nature of the transition process tself. (Perelman, 1980:394) Transition Period Issues: In his original paper, Lovins (1976) suggested that policy decisions would have to be made soon, either in favor of the hard or the soft path, because the time and resources absorbed by the hard path would "make the soft path less and less attainable" (Lovins, 1976:86). He did allow, however, the use of nonrenewable energy technologies during the transition period. The proper type and mix of energy for the "transition period" and beyond have been examined by various scholars and policy-makers challenging Lovins' claim for the "exclusiveness" of flow based technologies in the long term. Sweden opted for a national energy policy of keeping both 73 types of energy systems rather than prematurely closing options (Lonnroth et al., 1980). Some researchers argue that the two paths are not mutually exclusive (Bezdek et al., 1982: Barbour et al., 1982). A mix of large and small energy systems is necessary for the diverse types of requirements of society. They do warn, however, that "such a mix will require a deliberate effort to develop the largely untapped potentialities of smaller systems" (Barbour et al., 1982:74). Schnaiberg (1983) suggests that by allowing the use of some hard energy technologies during a transitional period, Lovins and other advocates of the soft energy path have been coopted by the hard path. Other Criticisms of the Soft Path Theory: Besides the early criticisms which focused on the precision of the figures used by Amory Lovins to estimate the need for a change to renewable energy based technologies (Inhaber, 1979: Nash, 1979; Edison Electric Institute, 1977), there have been criticisms of the theoretical relationships between technology, values, and social change. The argument which the Lovinses develop challenges the control of the economic and political organizations of society. However, it does not challenge the ideologies which support those structures. The Lovinses do not 74 challenge the legitimacy of science and technology as the foundation of economic development or the ideology of the free market. Yet these ideologies have supported the current hard path. The Lovinses propose redirecting those ideologies to support the soft path. Very little critical attention has focused on the micro- level implementers of the energy base changes at the community or the household levels. These are the primary actors proposed by the Lovinses to develop the soft path. Assumptions are made that households and communities will perceive it to their economic, political and personal advantages to withdraw from the current nonrenewable energy base of society and to develop an alternative energy base. The Lovinses assume that household members are aware of the current system vulnerability and have sufficient knowledge, economic, political resources and will to develop an alternative once governmental subsidies to the nonrenewable base are removed. In a previous section, I have discussed the perceptions of system vulnerability during an historical period when energy vulnerability was made quite salient in the 1970-1980 decade. Perceptions that an energy problem (e.g. system vulnerability) did exist were not widespread. The historical developments since that period have further 75 undermined perceptions of energy based system vulnerability. The Lovins' theory also assumes that individuals will want to develop an alternative system. The empirical evidence for this is sketchy. Although many households did adopt energy conservation measures without external subsidies, the energy base of the households (e.g. adoption of solar technologies or other renewable technologies) generally was not changed. Federal, state and community policies supportive of such a change seem to not have been very useful in stimulating the adoption of the solar technologies. The Lovins theory assumes that attitudes favorable to developing a new energy base are closely associated with implementing behavior. The social scientific literature is replete with examples where this is not so (Schuman and Johnson, 1976). (However, the relationship remains an empirical question in the case of solar energy technologies. This is examined in this project. CHAPTER IV THEORETICAL SPECIFICATION AND MODEL: SPP DEVELOPMENT Some of the criticisms of the Lovins' theory are based on lack of a specification of the conditions under which households and communities will participate in developing the soft path. There has been lack of attention to interest groups supportive or antagonistic to soft path developments and to potential conflicts. The relationship between the implementation of the renewable energy based technologies and their effects on attitudes and behaviors supportive of the soft path also has not been adequately explored. I begin by an explicit statement of Lovins' theory as a series of related propositions. Then I briefly describe a general natural resources vulnerabilities model. A specification and modification of Lovins' theory focusing on householders' use of solar energy technologies, is proposed. In turn, I describe the primary factors influencing the development of soft path preferences and their interrelationship with the soft path's restructuring of society. Finally, I present a model of soft path preferences development with empirically testable hypotheses. 76 77 LOVINS' THEORY OF SOFT ENERGY PATH SOCIAL CHANGE Lovins Ideas: Lovins' theoretical argument is more formally specified as follows: 1. Currently the world is on a destructive trajectory created by the characteristics of the energy base and attendant social structures of industrialized societies (i.e., the hard energy path). 2. The stock energy base has created systemic vulnerabilities due to the centralized, technocratically controlled and dangerous nature of the energy resources (e.g. uranium) of that energy system. 3. The nature of the energy system has created negative impacts on people in nations following the hard energy path. People of the industrialized nations experience alienation and lack of control over local community developments and resources because these are controlled by distant elites. The negative impacts also include:(a) chaos when the centralized energy system does not work appropriately, (b) international and national regional conflicts over control of energy resources, (c) greater social inequality 78 between generations, nations and classes as resources consumption differentials are exaccerbated and (d) potentially catastrophic dangers created by sabotage of nuclear facilities and materials, and by terrorists. 4. The negative effects of the hard energy path can be reduced by changing the energy base of society. 5. The energy base of society should be changed to a renewable energy base which is under the control of the energy end-users, e.g. the general population. The type of renewable energy technology used in this change should be determined by the work expected from the unit of energy (e.g. energy quality and end-use critera). Therefore, a variety of renewable energy technologies will be used. Most of the renewable energy technologies will be small, decentralized, natural resources conserving, and understandable to the user. 6. The decision to use the renewable energy technologies will be made noncoercively by the general population and communities. The population will be motivated to acquire 79 the renewable energy technologies because those technologies are more compatible with their values than are the centralized elite controlled hard energy technologies which currently constitute the hard energy path. Attitudes which are compatible with the soft energy path have developed in the American population. These include values of "thrift, simplicity, diversity, neighborliness, humility and craftmanship" (Lovins, 1977:57). They also include: (a) self-reliance, (b) natural resources conservation, (c) personal economic and noneconomic gains, «a (d) equity, and (e) social diversity. 7. The barriers created by governmental units and utility companies' subsidies of the hard energy path technologies make those technologies seem more economical than the soft energy technologies. The subsidies should be eliminated to allow the renewable energy technologies to show their competitiveness in the marketplace with the hard energy technologies. 7. The marketplace should provide the technological delivery system for renewable energy technologies. The market should however be locally based in order to provide economic multiplier effects to the community. This, plus a reduction in economic resource leakage to other places will 80 increase the employment and other economic resources available to the local population. Communities should also democratically and noncoercively participate in the diffusion of decentralized renewable energy technologies. The aggregation of households at the community level will benefit from the economic multiplier effects and the self-reliant control over the energy system. 8. There are several direct effects of the implementation of renewable energy technologies using the soft processes described above: (a) The effects on the local community will include greater self-reliance and economic development; greater democracy and social cohesion: greater social equity, thus lower stratification, and greater diversity. There will also be fewer socially catastrophic possibilities created by its energy system. (b) The direct effects on individuals and households include greater self-reliance and control, greater economic gains, greater access to the political processes of the community through democratic decision-making, and less alienation. (c) The direct effects on the natural environment create 8l more resources for communities, nations, and generations. They include more natural resources conservation and less pollution and other environmental problems. The indirect effects include the social structural transformations to the soft energy path as described by Lovins (1976, 1977). 9. Although the renewable energy technologies will eventually change the social structure, there are some measures to be taken during the transitional period. First, energy must be conserved to provide more resources and time for the transition. This step is economically beneficial for everyone. Second, a mix of nonrenewable and renewable energy technologies is allowed for the transition period until society is restructured creating fewer barriers for the renewable energy technologies. Eventually, there will be a complete change to a renewable energy base. 10. Once the renewable energy base has been acquired by the population, the indirect effects of the renewable 82 technologies will restructure society into the soft energy path even without changes in values and lifestyle. The world will then have more energy resources and will have pulled back from its trajectory toward a catastrophic brink. Suggested Additions: I have taken Lovins' basic proposal, added some variables to more clearly specify the theory, and then suggested some unspecified steps which exist between the change in the energy base of society and the social structural transformation of that society. Lovins starts from the assumption that the social system is currently in a very vulnerable situation but he does not develop the idea of perceptions of vulnerabilities or of many barriers in his theory. Schnaiberg (1983) is the first one to formally propose that knowledge of system vulnerability is an important part of the Lovins' theory. Since I agree with him, I have included the knowledge about system vulnerability and household experiences of that vulnerability in the specification of Lovins' ideas in Figures 2 and 3. The second addition which I have made to Lovins' thought is in more carefully specifying the resources which are needed 83 for household acquisition of renewable energy technologies. Not only must the "barriers" of policies and subsidies be considered, but other resources, such as personal network support and access to household socioeconomic resources should be examined. Increased resources and decreased barriers will reduce the cost of implementing values through the use of the technologies. Thirdly, I have suggested that not only is the adoption of technology affected by values, but technologies also affect the development of values. In this, I am agreeing with Veblen, Ogburn, Mumford, Stern and Aronson, and Mesthene. In fact, compatible attitudes and behaviors must proceed from these values to guide the use of the increased energy resources which are created through the use of renewable energy technologies, otherwise there will simply be a continuation of the hard energy path. Energy conservation is the initial expression of the values, followed by acquisition of solar energy technologies and behaviors which are also supportive of other natural resources conservation. These values will be expressed in the attitudes and behaviors captured with the soft path preferences scale. Fourth, although not formally proposing this variable in the 84 soft path preferences development model, it is imperative to consider the conflicting groups and consequences of a change to a renewable energy base. These are presented as obstacles to the implementation of social structural changes in Figure 2. SPECIFICATION OF NATURAL RESOURCES VULNERABILITY According to Lovins, household-implemented technical solutions will aggregate to lead directly to changes in the social structure of society (e.g. socio-cultural, economic and political impacts). Although Lovins' theory has focused on energy, its basic argument is relevant to any resource which is a basic need of the population such as air, water or food (Kinsley, 1984).8 The modified theory in its most general form is specified in Figure 2. Figure 2 Vulnerability Needs: The first variable driving the model is perceived vulnerability needs. System vulnerability will be perceived as an aggregate problem when the household members acknowledge the existence of a crisis in the resource base of their social system. This aggregate social system vulnerability will be perceived at either the national, regional, or community level depending on the salience given 85 to the crisis by the political (policies), economic (advertisement) and communication systems (mass media, personal, or other) which operate within the social environments of the household. This aggregate system vulnerability will also be experienced as household vulnerability. Economic signals of rapid price increases, for basic resources may be one stimulus to the perception of household and system vulnerability. Whether or not the household decision-makers change the household resource base will be influenced by the structure and composition of the household. The number of people dependent on the household, the stability of their relationship and their ages will affect resource use and demand elasticity. 86 >HHAHm mmomzommm q oo>fiooso mowcmsu moaomumoo meow>meon mcoepseom mcowusaom soumzm A. Hmssuossum ocm elllu Meowcnooh HmUM::OOH Hmwuom Hmwoom moosuwuu< mo mo coumwmco oeumsam> coeuoop< moossomom Hmsuxoucou oo>wooeoc 87 The interrelationships among these factors create the perception of vulnerability needs. The needs are weighed against the resources available to the household (e.g., possible substitute resources) and a decision is made about adopting a technical solution. The perceived household and societal vulnerability needs will directly affect the creation of consistent attitudes and behaviors as seen in Figure 2. Greater perceived system and household vulnerabilities are also associated with more supportive attitudes and behaviors. The available contextual resources affect the outcome of household decision-makers' weighing the economic and social costs of meeting the needs (e.g., resolving the disequilibrium of vulnerability). The contextual resources may be supportive, neutral, or in conflict with the perceived needs. The more resources available to the household, the less the cost of satisfying the needs and the greater the probability of adopting a technical solution. The contextual resources are composed of those which belong to the household, and those which are available in the community. The contextual resources which facilitate or impede the household's needs satisfaction include conflicts experienced in attempts to influence, acquire or use the household's resources. These processes may occur through interactions with extended family, friends, work associates, 88 neighbors and other acquaintances. They also are experienced with political systems, through national, regional, or community-based laws and policies, and with economic structures, such as lending, selling and service institutions. Adoption of Technical Solutions: Issues of economic cost, performance, reliability and risks of alternative technologies are assessed as part of the total costs and benefits to the household. If the assessment is generally positive, the household will resolve its vulnerability needs by acquiring a technical solution . These will further develop supportive attitudes and behaviors as barriers impeding such developments are reduced. Evaluation: For those households which adopt a technical solution to the perceived resource vulnerability, the experience of using the technologies will affect their evaluation of the technologies and therefore their attitudes and behaviors. Are their experiences with the technical solution consistent ‘with the original perception of needs? Was this solution implemented in ways consistent with a restructuring of society? 89 Consistent Attitudes and Behaviors: I have argued that technical changes are not sufficient. These must be associated with a package of consistent attitudes and behaviors which are supportive of the restructuring of society. This package directs the increased resources resulting from implementation of the technical solution. Social Structural Obstacles or Support: Although Lovins proposes that implementation of technological solutions will directly lead to social structural changes, this has been strongly questioned by some social scientists. Besides having a package of values to guide the results of the technical solution, those proposing to change the social structure of society encounter barriers created by present practices and control of societal resources as well as potential hostile forces of the dominant elites. This is where the walls created by "power without participation" are encountered. When these obstacles are overcome, then there will be social system changes. Social System Changes: The social system changes which will occur may be very different from those sought by the drivers of the changes. These changes will be the results of the interplay of 90 various forces: (1) experiences with the technological solutions: (2) the package of values (as expressed through attitudes and behaviors) which is reinforced by the technologies, and (3) interactions with elites, societal policies and practices, and with other people. The technical resolution of resource vulnerability needs also creates some social-structural impacts regardless of attitude and behavioral changes. These social system changes then provide feedback through pressure to reduce structural obstacles further changing attitudes and behaviors to become more consistent with the "new" social structure. The changes also encourage further positive evaluation and implementation of technical solutions consistent with the new structures. The feedback process will further decrease the vulnerability needs as presented in Figure 2. The general resource vulnerability model is specified to soft energy path structural changes, as follows. Specification to Energy Vulnerability: Between 1970 and 1980, household members perceived the existence of an energy crisis making the U.S. vulnerable to international economic and political forces. This energy 91 vulnerability was also experienced by households through increased cost of meeting basic heating and cooling needs. This vulnerability was felt more strongly by larger households in dwellings which consumed more energy. The contextual resources of households include their income and educational levels. Those having greater income and higher education have more knowledge of resources and disposable income to address the perceived energy vulnerability. The support which personal networks (family, work, friends, neighbors) provide in identifying the vulnerability and solutions is a contextual resource. Policy and legislative support for changing to a renewable energy base includes information distribution, financial incentives for acquisition of solar energy technologies, and laws requiring renewable technologies and conservation behaviors. Economic support includes lending institutions' willingness to provide loans for the acquisition of solar energy technologies, insurance coverage, and reasonable technology costs. The technical support includes repair services, solar builders and contractors, and solar stores. If the household has high income and education, support for solar technologies in its personal network, plus 92 governmental, business and technical support, then the perceived cost of acquiring solar energy technologies is reduced and the technologies are more favorably evaluated. The evaluations include characteristics of the technologies themselves such as their performance, reliability, and economic costs. For those who adopt the solar energy technologies, the assessment includes the experiences of owning the technologies. When the renewable energy technologies are implemented in a manner consistent with the theoretical ideas of soft path structural changes, this will have greater impacts on systemic changes. To be consistent with soft path changes, the solar energy technologies should be implemented in a participative fashion, based primarily on local community and neighborhood resources. Their operating problems should also be locally resolved. Passive solar systems for water and home heating and cooling meet the criteria of soft technology more completely than do active solar systems. Passive systems are appropriate to the end use of heating water or the home to a warm temperature. They can be made of locally available stones, plastic, black paint, etc. and use a renewable resource 93 (e.g. solar energy). The principles of heat absorption of the rocks, black paint, and the convection currents which are used to circulate the heat in the water or home are also easy to understand. The technology is under the control of the user for the user must participate in the activities of opening and closing windows, opening and closing apertures, etc. The technologies are also decentralized in the home of each user and self-reliant, for they don't require an external source of electricity. The acquisition process of passive solar energy technologies is more congruent with the theoretical characteristics of soft technologies than active solar energy systems as seen in Table 1. Table 1 The owners of passive technologies participate in the construction of the technologies and use local firms more often than active solar technologies owners. The latter predominantly use contractors and national firms. Passive solar owners do more of their own system design and installation, building them directly into their homes on site. Active solar systems tend to be added to the house as components or packages. These characteristics indicate that ‘the passive solar energy systems are more congruent with ILovins' notions of soft energy technologies and the SOLAR TECHNOLOGY ACQUISITION PROCESS 94 Table 1 I Active Active | Passive Passive System 1 System 2 | System 1 System 2 l Attached to n - 2435 | n - 923 House | | Built in 27 (868) 31 (153) 1 70 (767) 63 (183) I Added 70 (2240) 61 (301) | 24 (262) 26 (77) I Room added 3 (96) 9 (43) | 6 (70) ll (31) I Missing data* 5 (173) 76 (2880) | 7 (275) 28 (1083) Built | On site 15 (490) 34 (171) | so (359) 79 (222) I As component 28 (907) 33 (166) | 11 (114) 13 (37) I As package 54 (1733) 31 (149) | 8 (84) 7 (19) I Other 2 (70) 1 (11) l 2 (19) 1 (4) | Missing data* 5 (177) 85 (2880) l 8 (298) 29 (1092) Installation | Self 27 (864) 44 (221) I 42 (453) 51 (142) | Self and l (43) 1 (5) | 2 (24) 2 (7) contractor | l Contractor/ 72 (2303) 55 (272) | 56 (605) 47 (132) other I | Missing data* 5 (167) 76 (2879) I 21 (292) 29 (1093) Design Active Small local firm 38 (1025) 37 (151) - - - - Small national 21 (559) 21 (86) firm - - - - Large national 31 (853) 26 (105) firm - - - - Foreign/other 10 (283) 16 (65) - - - - Missing data* 6 (166) 86 (2479) - - - - Design-Passive Self Acquired plans Professional/ other Missing data* 51 (54a) 56 (159) 3 (35) 4 (10) 46 (500) 40 (113) 8 (291) 29 (1092) * Missing data include those respondents for whom the question is not applicable. For example, passive solar owners would be classified as missing if the acquisition process is applicable only to active systems. The non-missing values should add to 100%. 95 development of soft path preferences. development of the SPP. Holdren et al. (1980) support this in concluding We think it likely that, assuming sensible materials choices, the use of passive solar design in architecture will produce smaller environmental impacts than those resulting from supplying an equivalent amount of energy from any of the 'active' technologies. (Holdren et al., 1980:249)- Nebraska homebuilders found passive solar systems easier and more cost effective to build, and easier to explain to the consumer. They reported building or seriously considered building more passive systems than active systems. Loan officers were very positive about making loans for passive solar homes (Combs, 1983). Vine (l980a) reports that members of solar energy associations used more passive systems or mixed systems (active and passive) than active systems. Vine (l980a) examined the differences between active and passive owners concluding 96 Owners of active solar systems are more likely to be wealthier, married, and homeowners than owners of passive and hybrid systems. Not surprisingly, the hybrid group is usually located between the other two groups in their socioeconomic characteristics. The three groups do not statistically differ in sex, age, occupation, education, and ideology. (Vine, 1980a:156). Federal programs researchers (Farhar-Pilgrim and Unseld, 1982: Vine, 1980a) have examined primarily active solar sytem owners. It is theoretically important to conserve the distinction between these two types. Finally, the households which experience energy vulnerabilities and resolve them by changing to renewable energy base are hypothesized to have different attitudes and behaviors than the households which do not change their energy base. These preferences will affect the type of path changes which will develop as seen in Figure 3. Figure 3 The complete theory of the transition to the soft energy path moves from perceptions at the household level to social structural impacts at the societal level. Much of the household level decision-making process is assumed by Lovins, who focuses primarily on the social structural impacts. 97 Ikommzm boom mIH OH ZOHHHmzeousoo _ tendon \\\\Mwwmeomoso moemoflocnoob mmwmoeoccuoe ommAlllllcumo u omAIIIIIsmHom moAIlIIIIIIsmHom coHumsHm>m mo Wash t:£ei-athI/xz )///moos30mmm ,f/fi/if- -: Hmsuxoucou to>eoosoo 98 My study, however, examines the household level dynamics assumed to be occurring to produce the changes to the soft path. The model I am examining is expressed in Figure 4, showing households which have not and households which hays changed to a renewable energy base. MODEL SPECIFICATION An examination of the soft path change theory requires community level impact measures (Morrison and Lodwick, 1981). However, it is possible to examine part of the theory using household data. Figure 4 shows two models of part of the theory which I am analyzing. I will focus on the development of the soft path preferences which are supportive of soft path changes. I am omitting the social structural impacts and their feedback influences which have been shown in the previous section. The models are specified for two different groups, renewable energy system owners and nonrenewable system owners. Figure 4 Since the nonsolar households do not have experience with ownership of renewable energy technologies, their opinions of solar technologies and the contextual external resources, are necessarily more abstract. Energy issues are more invisible to them. The model allows me to examine how this 99 (Nonsolar) Perceived——#*’”"'———F Contextual Resources Soft Evaluation of Path /Solar Technologies _>Preferences Perceived/ Energy Vulnerability (Solar) Perceived Contextual Resources Type of Evaluation of Soft Solar———————+ Solar >Path Technologies Technologies Preferences Perceived Energy Vulnerability Figure 4: DEVELOPMENT OF SOFT PATH PREFERENCES 100 influences the process of soft path preferences development for the nonrenewable energy system owners. The model is also expanded to include the experiences of the renewable energy technology owners. Their perceptions of contextual resources, solar technology type and evaluation are disciplined by experiences. Their household energy vulnerability needs are based on the results of already having a behavioral resolution to perceived disequilibrium. Therefore energy issues are more visible to them. The soft path preferences measures are the same for both the renewable and nonrenewable system owners. The models allow me to examine how the process of SPP development is affected by perceptions of energy vulnerability needs, contextual resources for solar energy technology acquisition, evaluation of the technologies with and without the experience of owning the technologies and the influence of the characteristics of the active and passive solar technologies. The model's variables have been specified above. The operationalization of the variables is discussed in Chapter VI. The model will be examined empirically using a U.S. national 101 probability sample of nonrenewable energy system homeowners using a stock energy base for their heating and cooling needs and a purposive sample of renewable energy technology owners who use primarily9 solar energy technologies for home heating and cooling. The solar system sample will be further subdivided into (1) active and passive solar technology owners, and (2) different ownership-length samples. There are three basic questions addressed by this project: 1. Is there a greater degree of soft path preferences among those who have changed the energy base of their household compared to those who have not changed their energy base? Operationally: H1:There is a greater amount of SPP among solar homeowners than among nonsolar homeowners. I; the values and attitudes of the homeowners who have changed their energy base are more consistent with the soft path, then that is an indication of potential support for a transition to the soft energy path. Both the renewable energy base technologies and soft path preferences are necessary for social structural changes of the soft path. 102 2. What has the greatest influence on the development of SPP of the solar and nonsolar homeowners? Qperationally: H2:The solar technologies are more strongly associated with the soft path preferences than the perception of energy vulnerability and the availability of contextual resources. If this hypothesis is correct, then I conclude that the use of solar energy technologies is more critical to soft path changes than further emphasis on the dimensions of soft path support. 3. Does the type of solar technology and the length of ownership of the technologies have an influence on the development of SPP? Hypothesis three will compare owners of the different types of technologies while hypothesis four examines the whole solar sample. Operationally: H3:Passive solar energy technology owners have more SPP than active solar energy technology owners. H4:Length of ownership of solar technologies is positively associated with degree of SPP. If hypothesis three is correct, then there is an indication of further consistency in the relationship of soft renewable energy technologies and soft path changes. Support for hypothesis four suggests that the influence of the renewable energy technologies is developmental over time. .Alternatively, solar owners may simply experience short term enthusiasm and then a return to the hard path rather than 103 providing a solid foundation for soft path changes. To the extent that these hypotheses are supported empirically, then there is some basis for claiming that a change in the U.S. residential energy base is providing a supportive cultural context for the development of a restructured society. However, it is important to recognize that only 20% of the energy consumed in the U.S. in 1980's was used by home heating and cooling (Green et al., 1984). This project does pg; address the very critical linkage between the positive cultural context and the social structural impacts of changing the societal energy base. It also does not address the role of elites in controlling such a change except as experienced indirectly through householder's perceptions of contextual resources and conflicts. CHAPTER V RESEARCH AND SAMPLING DESIGN AND SAMPLE CHARACTERISTICS In this chapter, I describe the research design and then the samples. The sampling design, data gathering methods and instrument development are then discussed. Finally I compare some of the characteristics of the solar energy technology owners with those of the nonsolar energy technology owners. RESEARCH DESIGN This study uses data gathered from a U.S. national probability sample of homeowners who did 325 own solar technologies and a purposive sample of homeowners who owned them in 1980. Face to face interviews were conducted with the nonsolar respondents by fieldworkers of the Gallup Organization Inc., while a mailed questionnaire was sent to the solar respondents by SERI (the Solar Energy Research Institute) under the direction of Dr. Barbara Farhar- Pilgrim. The solar homeowners sample was then subdivided into passive (9%, n-326) and active (60%, n=2278) solar technology owners. These were self defined solar technology owners who claimed to own only passive or passive hybrid 104 105 or only active systems. The 32% (n-1205) who claimed to own both passive and active solar technologies were not used for further analysis. The active and passive technology owners are theoretically representative of two approaches to changing the energy base of society. Later I will argue that the passive solar technologies are more "soft” than the active solar technologies. Finally, the solar sample was divided into four subgroups based on length of technology ownership in 1980: (1) ten years or more (n-116), (2) 5-9.99 years (n-285), (3) 1-4.99 years (n-2728), and (4) less than one year (n-680). A soft path preferences (SPP) scale was developed. The solar and nonsolar homeowners, the active and passive solar technology owners, and length of ownership samples' SPP scale scores are compared in order to determine which group had preferences more congruent with the development of soft path social change. Finally, the variables influencing the development of their soft path preferences were compared for the solar and nonsolar homeowners and the active and passive solar technology owners using the SPP development model proposed 106 previously. These comparisons provided the data used in examining the hypotheses of this study. DATA GATHERING PROCESSES Sampling Design: The population of the nonsolar sample was defined as "all owner-occupied year-round housing units in the 48 contiguous states, including mobile homes, but excluding town houses, condominiums, and military bases" (Farhar-Pilgrim and Unseld, 1982:34). This is because people living in those dwellings did not have decision-making power over the external appearance of the building. Those living in attached houses who did have such power, as ascertained through a screening question, were included in the sample. The number of households in the U.S. in 1978, which was the latest available figure at the time the surveys were conducted, was 50,283,000. This figure was updated to include information from the 1980 Census. The Gallup Organization then used the relationships between the population and the number of owner-occupied housing, and between the population and the number of households to estimate that there were 52.2 million year-round owner- occupied housing units in the contiguous states - excluding (Alaska and Hawaii - in the fall of 1980. 'The Gallup Organization has a national probability sampling 107 frame of 362 interviewing areas. All the areas were included in the sample design. The sampling method included replicated, probability sampling to the block level for the urban areas and to segments of townships for rural areas. The sample size of approximately 2,000 was determined by analytical needs for having subgroups of sufficient size for predetermined levels of accuracy. The estimated standard errors for the questions based on sample variability ranged from .009 to .018. "The 95 percent confidence interval for estimates with standard errors in the middle of this range would be about 2.6 or 2.7 percentage points, plus or minus" for the whole sample (Farhar-Pilgrim and Unseld, 1982:36). The final useable sample size of the nonsolar homeowners was 2,023. In 1980, it was impossible to identify the solar homeowners' universe. ”'Solar homeowners' were defined as owners of homes employing any type of residential renewable energy systems" (Farhar-Pilgrim and Unseld, 1982:37). Those who had wood-burning stoves or fireplaces as their only renewable energy systems were not counted, however. ,A list of 6,911 names of probable solar homeowners was compiled from the National Solar Heating and Cooling ‘Lnformation Center, state energy offices, from other researchers, from nineteen published directories of solar 108 owners and also from direct responses of individuals to three press releases in over 1,500 publications throughout the U.S. About 40% of the solar respondents volunteered to participate in the survey, so it is impossible to determine how well the sample represented average solar homeowners of that period. Of the names on the original mailing list, 767 were identifed as nonsolar homeowners or the addresses were inaccurate. Therefore the sampling frame was reduced to 6,144. After several follow-up procedures, 3,809 completed survey instruments were returned for a response rate of 62%. Of the 2,335 persons who did not complete a questionnaire, 2,157 did not respond at all. The following reasons for nonparticipation were given by 96 individuals who responded: (l) 44 gave no explanation, (2) 21 showed "displeasure with the survey", (3) 9 said they were having problems with their solar systems, (4) 6 were unhappy with the Federal government, (5) 5 said the survey was a waste of time or tax money, (6) 4 said the survey was "too personal", (7) 5 claimed to be "too busy” and (8) 2 were physically disabled (Farhar-Pilgrim and Unseld, 1982:38). Other categories of incomplete questionnaires included: (1) 23 invalid or incomplete responses, (2) 18 were late e.g. 109 past the end of December, 1980, (3) 3 had return address labels removed so were unidentifiable, and (4) 33 blank questionnaires were returned (Farhar-Pilgrim and Unseld, 1982:39). Data Gathering Methods: The interviewers for the nonsolar personal interviewing process were trained by Gallup Organization "through the use of written instructional materials, a tape-recorded summary of key information and telephone briefings with each interviewer" (Farhar-Pilgrim and Unseld, 1982:36). Fieldwork was conducted from October 16, 1980 to November 24, 1980. The work was interrupted from the weekend prior to the national election held on November 4, to the Wednesday after Election Day. The average interview length was 56 minutes. Gallup reported that during the pretest the survey had created a lot of interest and people had many questions. So the interviewers left pamphlets and information on how to get more information about solar energy with the respondents after the interviews (Farhar-Pilgrim and Unseld, 1982:36). Thirty percent of the completed interviews were validated ‘with follow-up telephone calls or through the mail. "The level of refusal encountered in the survey was llO approximately the same as that which Gallup has encountered in similar work" (Farhar-Pilgrim and Unseld, 1982:37). The potential solar respondents were sent a mailed questionnaire with a cover letter and postage-paid return envelope. After a two week period, a follow-up postcard was sent, then a second postcard after two more weeks. Finally, a second questionnaire and cover letter were sent to all who had not responded two weeks later. This process was completed by October, 1980. The responses were accepted through the end of December. Marketing Management Concepts Inc. was subcontracted by SERI to edit and code the questionnaires. The return address labels on the envelopes were destroyed to assure the anonymity of the respondents. The marketing firm then provided SERI with a data tape, codebook and frequency distributions of all the variables (Farhar-Pilgrim and Unseld, 1982:39). Obtaining the Data Tapes: After hearing of the SERI study, I decided to investigate its potential as a data base for an analysis of soft path preferences. I met with Dr. Farhar-Pilgrim in the summer of 1981. She assured me that there would be no problems in my using the data and suggested my contacting SERI directly. I then attempted to obtain a copy of the data from SERI, which 111 had undergone reorganization by 1982. Although the nonsolar homeowners data were readily available, the solar homeowners' data were missing from SERI. Dr. Farhar-Pilgrim was able to provide me with another copy of the data tape with the proviso that I not publish the geographic locations of solar homeowners on a state by state or zip code basis, for that could be an "invasion of respondent privacy" in states where little solar technology existed (Personal communication, December 14, 1982). I agreed to that. Also, the identification numbers were stripped from the data, again assuring anonymity. When I received the codebooks and data tapes, the nonsolar homeowner data and codebook matched perfectly. However, that was not true of the solar data set. Therefore, I reconstructed the solar codebook from the structure of the data set. The accuracy of the reconstruction was validated with frequency runs compared to those reported in Farhar- Pilgrim and Unseld (1982). Because I did not have access to the original questionnaires, I was not able to correct the ten cases with miscoded variables in the solar data. These were recoded into "missing variables.” 112 Instrument Design: Various techniques were used to gather information to design the research instruments. Focused interviews were conducted with solar and nonsolar homeowners, solar leaders and community leaders of four communities known to be pro-solar throughout the U.S. - Davis, California: the San Luis Valley of Colorado: Carbondale, Illinois, and New York City. The tape recorded open-ended interviews were content analyzed for the ideas and concerns of people who had thought about and/or adopted solar energy technologies (Farhar-Pilgrim and Unseld, 1982:33). A six-month study was conducted of nine solar homeowners in Tucson, Arizona to further explore the relationship between "values and life-style preferences in the solar adoption decision" (Farhar-Pilgrim and Unseld, 1982:34). Members of the survey design team also met with Department of Energy officials, and "representatives of large and small firms manufacturing solar equipment, builders of solar homes, and public interest groups" (Farhar-Pilgrim and Unseld, 1982:33). These sources of information were then used to design the nonsolar interview schedule and the survey instrument for the solar homeowners. The survey instrument and "study plan were reviewed and approved by the Energy Information 113 Administration and the Office of Management and Budget" (Farhar-Pilgrim and Unseld, 1982:34). The survey instruments used with both samples were identical except for questions which were inappropriate because both did not own solar technologies. The solar survey instrument had many questions about specific experiences with solar technologies including operating, engineering, and economic factors. Attitudinal, behavioral and lifestyle preference questions were basically identical for both samples. Questions concerning the potential advantages and problems with solar technologies were asked retrospectively of the solar homeowners, e.g. "when you were thinking about using solar energy in your home". . . or ”how important this advantage actually was to you in making your decision . . ." They were asked hypothetically of the nonsolar homeowners, e.g. "if you were thinking about using solar energy in your home " and ". . . how important it [possible advantages] would be for you in making . . . a decision." Copies of the instruments are reproduced in Farhar-Pilgrim and Unseld (1982) Appendixes B and C. Most of the questions were structured but there were a few open-ended questions. .As seen in Appendix B of this document, the Likert format was used quite extensively in the questions of particular interest to this project. 114 SAMPLES' CHARACTERISTICS The solar and nonsolar samples' characteristics include house and fuel types, household composition, socioeconomic status indicators (household income, occupation of head of household, education of the respondent), and other demographic information such as age, race, gender, and geographic region. Housing and Fuel Characteristics: Housing and fuel characteristics summarized in Table 2 indicate current energy type. Table 2 Both the solar and nonsolar households sampled predominantly live in single dwellings (95% and 90%). Less than 10% live in multiple housing structures or in mobile homes. The greatest difference, but nevertheless a small one, is seen in the percentage of the samples living in multiple housing units (3% solar to 7% nonsolar). The main fuel types used by the two samples is quite different. Most of the nonsolar households used natural gas as their primary heating fuel (56%). Fuel oil, which was controversial in 1980 due to the oil embargo, was the ,primary fuel in 19% of the households. Electricity was used i:113% of the households, followed by lesser use of propane, 'wood and other fuels. 115 Table 2 HOUSING CHARACTERISTICS AND PRIMARY HEATING FUEL Solar Nonsolar % (n) % (n) HOUSING TYPE Single dwelling 95 (3565) 90 (1819) Multiple dwelling 3 (98) 7 (136) Mobile 1 (33) 3 (58) Other 1 (47) 1 (10) Missing data* 2 (66) O (0) Total 100 (3809) 101 (2023) MAIN FUEL TYPE Solar 18 (650) 0 (0) Fuel oil 16 (580) 19 (368) Electricity 32 (1181) 13 (249) Propane 2. (81) 4 (72) Natural gas 12 (426) 56 (1101) Wood 10 (383) 8 (161) Other 10 (365) 1 (17) Missing data* 4 (143) 3 (55) Total 100 (3809) 101 (2023) *Non-missing percentages based on those who answered. 116 On the other hand, the modal solar households used electricity (32%) as primary fuel while the second largest category of households claimed solar (18%) as the main heating source. Fuel oil was used by only three percent fewer solar than nonsolar households (16%). Both samples showed about the same percentage using wood as the primary heating system, though the solar sample has a slightly higher percentage of users (10% compared to 8%). The other point that Table 2 makes is that the solar sample used a more diversified array of fuel types when compared to the nonsolar. Table 3 presents the average cost of household energy use in the summer and winter of 1980 for the two samples. Table 3 On the average, the nonsolar households paid $103.00 per month for household heating, cooling, electricity, and hot water. The range varied from $4.00 to $938.00 with some reporting winter heating costs as high as $1,300! However, 75% of the households paid an average of $120.00 or less per month. The solar sample reported an average monthly expenditure of $93.54 for household energy. The range was $3.00 - $5,000.00 with 75% of the households paying $100.00 or less per month. 117 Table 3 AVERAGE ENERGY COSTS (in 1980 dollars) * l | Quartiles Summer | Winter | Monthly I I I | SOLAR | | I | lst 1 - 35 | 3 - 45 | 3 - 43 I | 2nd 36 - 50 | 46 - 70 | 44 - 66 I | 3rd 51 - 80 | 71 - 116 | 67 - 100 I I 4th 81 - 6,000 | 117 - 8,000 | 101 - 5,000 Missing data 9% - (347) | 12% - (448) | 13% - (477) Tota1 (3809) | (3809) | (3809) X $82 | $109 I $94 s.d. $358 | $280 | $209 Range $1 - $6,000 | $3 - $8,000 | $3 - $5,000 NONSOLAR | | | I lst 4 - 39 | 4 - 70 | 4 - 60 l | 2nd 40 - 60 | 71 - 100 | 61 - 85 l I 3rd 61 - 98 | 104 - 150 | 86 - 120 l I 4th 100 - 900 | 151 - 1300 l 122 - 938 Missing data 2% (45) | 3% (62) | 3% (68) Total (2023) | (2023) | (2023) X $77 | $129 | $103 s.d. $ 61 | $108 | $74 Range $4 - $900 I $4 - $1,300 | $4 - $938 * Conservative estimate based on averaging summer and winter energy costs. 118 Not only are the physical demands of the housing structure and energy costs important, but also the social composition and stability of the household. Table 4 describes the household composition of the homeowners. Table 4 The modal family structure is that of a two parent family with children, in both samples. The next most common household structure is married couples with no children (31% of the solar and 28% of the nonsolar samples). The greatest difference between the two samples is in the larger percentages of single adults and single parents in the nonsolar sample. Sixteen percent of those respondents were adults living alone, while 10% were single parents with predominantnly one or two children. Less than 10% of the solar sample fit these categories. These patterns may reflect the older age of the solar sample (see Table 8). The average number of children was very similar for both samples with 1.305 in the solar and 1.137 in the nonsolar households. Migration plans of households may affect the willingness of the homeowners to invest in renewable energy technologies. These plans were determined with the question, "How likely are you to move within the next three years?" Household Composition 119 Table 4 Solar Nonsolar % (n) % (n) FAMILY TYPE Two parent family 49 (1865) 42 (852) Married couple 31 (1168) 28 (569) Extended family 4 (165) 4 (82) One adult 1 (45) 16 (326) Single parent 5 (180) 10 (194) Other 7 (257) - - Missing data 3 (129) - - Total 100 (3809) 100 (2023) NUMBER OF CHILDREN 0 40 (1521) 45 (913) 1 17 (664) 19 (389) 2 23 (876) 20 (409) 3 10 (389) 11 (215) 4 4 (134) 3 (68) 5+ 3 (95) 1 (29) Missing data 3 (130) - - Total 100 (3809) 99 (2023) X 1.305 1.137 120 The nonsolar and solar samples were very similar. About 75% claimed that it was "very unlikely" or "unlikely" that the household would move during that period and about 15% claimed that it was "very likely" or "likely" that they would move (Appendix A, Table 24). Socioeconomic Characteristics: The economic, occupational, and educational characteristics of the samples provide indicators of the households' resources for implementing their preferences. The household income provides disposable income for acquiring solar energy technologies. The occupation of the head of household suggests the types of technical skills and knowledge which the householders have, as well as prestige. Table 5 As expected, the nonsolar sample generally had a lower income level than the solar sample. For example, 12% of the nonsolar sample fell below the 1979 Federal Poverty level guidelines for a family of three (U.S. Dept. of Commerce, 1983) while only 2% of the solar households was in that income category. 121 Table 5 SOCIOECONOMIC CHARACTERISTICS OF HOMEOWNERS Solar Nonsolar % (n) % (n) EDUCATION Graduate school 39 (1451) 9 (172) College 24 (890) 13 (252) Some college 19 (688) 18 (369) Trade or technical 6 (232) 6 (119) High school 10 (360) 32 (639) Less that high school 3 (97) 23 (452) Missing data 2 (91) 1 (20) OCCUPATIONS Professional 40 (1491) 16 (307) Manager,executive 15 (557) 9 (186) Business owners 7 (254) 6 (113) Skilled workers 11 (382) 17 (333) Retired 14 (523) 19 (383) Other 4 (156) 0 (1) Service worker 2 (79) 4 (88) Manufacturer's rep. 2 (59) 2 (34) Semi-skilled trade 1 (46) 10 (196) Sales .5 (17) l (24) Clerical, office 2 (59) 5 (99) Farm owner, mngr. 1 (41) 4 (70) Laborers .6 (21) 5 (98) Homemaker, .6 (19) 3 (52) Student Missing data 3 (105) 2 (39) INCOME Greater than $55,000 12 (423) 4 (74) $45,000 - $54,999 10 (341) 4 (75) $35,000 - $44,999 17 (617) 10 (186) $25,000 - $34,999 24 (853) 16 (302) $20,000 - $24,999 14 (510) 16 (287) $15,000 - $19,999 10 (366) 15 (273) $12,000 - $14,999 5 (175) 9 (169) $10,000 - $11,999 3 (102) 7 (125) $7,000 - $ 9,999 2 (87) 8 (140) Less than $6,999 2 (86) 12 (217) Missing data 7 (249) 9 (175) Total (3809) (2023) 122 When the samples were divided into quartiles by income, the quartile income values were lower for the nonsolar sample. The solar group had about $10,000.00 more in each quartile than the nonsolar sample. For example, the lowest quartile in the nonsolar sample included those making up to $11,999 while in the solar group, the lowest quartile went up to $19,999. The third quartile's upper range was $34,999 among the nonsolar households and $44,999 among the solar homeowners. Thus it is clear that the solar sample has access to greater economic resources. In this study, professionals are over-represented in the solar sample when compared to the nonsolar sample (40% to 16%). Managers and executives are also somewhat overrepresented in the solar sample with 15% compared to 9%. Business owners are approximately equally represented in the two groups with about 6% of the respondents being in this occupational category. Table 6 Table 7 Although the skilled trades do not rank in the top income generating groups, they may have more abilities to work with technologies. Although skilled tradespersons represent one of the larger groups within the solar sample, it is underrepresented when compared to the nonsolar sample (11% compared to 17%). This is also true of the semi-skilled 123 Table 6 INCOME AND OCCUPATIONS - SOLAR Very Upper High High Middle Middle Low Poverty ($45,000+) ($35,000- ($20,000- ($12,000- ($7,000- (Under Row Occupations 44,999) 34,999) 19,999) 11,999 $6,999) Total Professional 28 22 38 9 2 1 41 (401) (308) (548) (128) (29) (13) (1427) Manager/ 34 25 I 34 6 l l 15 executive (182) (132) (181) (31) (7) (4) (537) Own business 21 15 37 17 7 3 7 (51) (37) (89) (41) (17) (6) (241) Skilled 6 9 49 27 7 2 10 worker (22) (33) (175) (97) (25) (6) (358) Retired 7 8 33 28 17 8 14 (32) (37) (159) (134) (80) (36) (478) Other 15 12 41 20 7 5 4 (22) (18) (61) (29) (ll) (7) (148) Service 14 10 45 27 3 3 2 worker (10) (7) (33) (20) (2) (2) (74) Manufact. rep 29 14 48 7 2 0 2 (17) (8) (28) (4) (l) (0) (58) Semi-skilled 14 16 49 16 2 2 1 worker (6) (7) (21) (7) (l) (l) (43) Sales 0 18 53 12 12 6 l (0) (3) (9) (2) (2) (1) (l7) Clerical/ 0 16 47 23 14 0 2 office (0) (9) (27) (13) (8) (0) (57) Farm owner/ 16 13 26 32 5 8 l manager (6) (5) (10) (12) (2) (3) (38) Farm labor 0 0 33 0 33 33 0 (0) (0) (1) (0) (l) (1) (3) Laborer 24 12 24 41 0 0 1 (4) (2) (4) (7) (0) (0) (l7) Housewife 23 8 16 31 8 l6 0 (3) (1) (2) (4) (1) (2) (13) Full-time 0 50 0 0 0 50 0 student (0) (1) (0) (O) (0) (1) (2) Total 22 17 38 15 5 2 100 (756) (608) (1348) (529) (187) (83) (3511) Missing data 298 Grand Total 3809 INCOME AND OCCUPATIONS - NONSOLAR 124 Table 7 Very Upper High High Middle Middle Low Poverty (545,000+) ($35,000- ($20,000- ($12,000- ($7,000- (Under Row Occupations 44,999) 34,999) 19,999) 11,999 $6,999) Total Professional 19 20 45 10 5 2 15 (53) (56) (125) (28) (14) (5) (281) Manager/ 21 19 42 15 4 1 9 executive (35) (32) (71) (25) (6) (1) (170) Own business 11 24 34 20 5 5 5 (11) (24) (34) (20) (5) (5) (99) Skilled 4 10 45 28 8 5 17 worker (11) (32) (141) (89) (26) (14) (313) Retired 2 1 7 23 36 31 19 (7) (4) (24) (78) (125) (106) (344) Other 0 0 100 0 0 0 0 (0) (0) (1) (0) (0) (0) (1) Service 2 4 39 26 8 20 5 worker (2) (3) (33) (22) (7) (17) (84) Manufact. rep 37 3 33 17 7 3 2 (11) (1) (10) (5) (2) (l) (30) Semi-skilled 1 4 35 36 15 9 10 worker (2) (7) (63) (65) (26) (16) (179) Sales 0 15 35 30 15 5 1 (0) (3) (7) (6) (3) (l) (20) Clerical/ 2 5 35 40 13 4 5 office (2) (5) (32) (37) (12) (4) (92) Farm owner/ 17 14 16 28 14 11 4 manager (11) (9) (10) (18) (9) (7) (64) Farm labor 0 0 10 40 20 30 l (0) (0) (1) (4) (2) (3) (10) Laborer 0 8 33 35 17 7 S (0) (7) (28) (30) (15) (6) (86) Housewife 0 0 3 15 18 64 2 (0) (0) (1) (5) (7) (25) (39) Full-time 0 0 0 43 29 29 0 student (0) (0) (0) (3) (2) (2) (7) Refused 0 ll 26 21 21 21 1 (0) (2) (5) (4) (4) (4) (19) No answer 40 10 30 20 0 0 l (4) (1) (3) (2) (0) (0) (10) Column Total 8 10 32 24 14 12 100 (149) (186) (589) (442) (265) (217) (1848) Missing data 175 Total 2023 125 professions. Although these professions are represented in 10% of the nonsolar sample, they are only 1% of the solar sample. Other occupations which have low income are not very strongly represented in the solar sample as expected. Examples of these are sales, clerical, farm managers, and laborers. Homemakers and full-time students are also underrepresented in this group when compared to the nonsolar sample. There is some underrepresentation of retired people who are solar owners when compared to the percentage found in the nonsolar group (14% to 19%). However, the solar retirees have higher incomes than those in the nonsolar sample (Tables 6 and 7). Education is an important resource, for differing levels of education expose individuals to different types of information and help develop skills in seeking new information. While 55% of the nonsolar sample have a high school education or less, the solar sample is composed predominantly of those with a college level education or higher (63%). In fact almost 40% of the solar sample has graduate level education while the nonsolar homeowners have an equivalent percentage (38%) in trade or technical schools 126 and high school (Table 5). Other Demographic Characteristics: Other demographic characteristics include age, race and gender. Table 8 In these data, 21% of the nonsolar homeowners are below 34 years of age and 20% are 65 or older. The solar sample shows a distribution which is slightly smaller in these age groups - 16% under 34 and 16% are 65 or older. The curvilinear relationship of age and solar system ownership is apparent as 68% of the solar respondents are in the 35 to 64 age groups, while only 59% of the nonsolar respondents are in these categories. The mean age, however, for the two groups is very similar - 43 years for solar and 44 for the nonsolar. The ethnic and racial composition of the U.S. population was well reflected in the nonsolar sample. The predominant group was Caucasian (88%) as was true in the solar sample (85%). The second largest group was Blacks, representing 8% of the sample. This ethnic group is significantly underrepresented among solar homeowners (.4%). This is also an underrepresentation of the Blacks in the U.S. population DEMOGRAPHIC 127 Table 8 CHARACTERISTICS OF HOMEOWNERS Solar Nonsolar % (n) % (n) AGE 24 or less .4 (l4) 4 (76) 25 - 34 16 (608) 17 (345) 35 - 44 26 (970) 20 (403) 45 - 54 22 (799) 21 (414) 55 - 64 20 (716) 18 (364) 65 - 74 13 (467) 14 (270) 75 or more 3 (94) 6 (127) Missing data 4 (141) l (24) Total 100.4 (3809) 100 (2023) RACE Caucasian 85 (3110) 88 (1761) Asian 13 (467) 1 (16) Black .4 (l6) 8 (165) Hispanic 1 (40) 2 (44) American Indian 1 (22) 1 (11) Missing data 4 (154) l (26) Total 100.4 (3809) 100 (2023) GENDER Male not available 54 (1072) Female --- ------ 46 (910) Missing data --- ------ 2 (41) Total 100 (2023) 128 (U.S. Dept. of Commerce, 1983). The geographic distribution of the two samples does not explain this for both have about 27% from the South. The high socioeconomic level of the solar homeowners is a better explanation. This is probably the reason why the Hispanic and American Indian populations were underrepresented. Many of these ethnic groups have a high percentage of their members in poverty and are renters rather than homeowners. The Asian/Pacific Islanders (12%), however, are overrepresented. This may reflect the overrepresentation of the Western region (39%) in the solar sample - especially since it includes Hawaii and California, two states which have both a high percentage of people of Asian descent as well as a preponderance of solar technologies. Although the gender of the respondents is important in influencing attitudes towards energy and environment, this information is not available for the solar respondents. The nonsolar sample, however, is composed of 53% males and 45% females. Women have been identified as having more favorable environmental attitudes, stronger energy conservation behavior (Farhar et al., 1980), and as being more strongly against nuclear power and pro-solar than men 129 (McStay and Dunlap, 1983: Stout-Wiegand and Trent, 1983; Reed and Wilkes, 1980: and Brody, 1984). Brody (1984) examines two hypotheses in his work - one which suggests that the reason for women's lack of support for nuclear power is due to their marginality in the social structure, and the other, that it is due to their perception of the safety risks associated with the systems. Brody concluded that although men and women viewed the energy crisis as serious, women were less supportive of nuclear power as a way of resolving the crisis because of their concerns with the risks. Reed and Wilkes (1980) also concluded that women were more strongly against nuclear power development than men because of their marginality in the social structure. This is somewhat supported by Stout-Wiegand and Trent's (1983) findings that women were more strongly against economic and energy developments than men because of the negative social impacts of boomtown conditions. Women have engaged ”in behaviors such as cutting down on driving, saving newspapers for recycling, and avoiding environmentally damaging products (of all types) more often than do men . . . reflecting a genuinely stronger commitment to a 'pro-environmental' lifestyle" (McStay and Dunlap, 1983:297). Men on the other hand, do engage in more 130 political actions. These findings suggest that women will be more favorable to the development of solar energy technologies than men. However, Farhar-Pilgrim and Unseld (1982) found that men were more favorably inclined toward solar energy systems than women. They also found that women are more likely to perceive the energy situation as getting worse in the‘ future. Unfortunately this issue cannot be pursued further with these data. Rggional Characteristics: Garrett-Price et al. (1980) report that the West shows the most immediate promise for solar space heating as an economic option. This may be reflected in the larger percent of the solar sample in the West (39%) compared to the nonsolar sample (14%) (Table 9). Table 9 Compared to the nonsolar sample, there is an under- representation of the solar respondents in the Midwest (8% to 31%). Surprisingly, there is an equivalent representation of solar and nonsolar households in the East (27%) and very similar representation in the South (26% solar to 28% nonsolar). 131 .oofim Oppose .som>oz .mcfisoxz .osmucoz .mxmch .flflmsom .msmzmaoo .ouoxmo :uuoz "Moaomcoc you ooucmmmhoou uos mousum as .muoxmo nusom .soc>mz .mcmemssoq ”umaom no“ noncommuomu no: moumum k Ammomo cod Amommo oos Hmuoe on 0 Amy 0 sumo osemmflz Ass.ez.x<.zm.z.o meson xmx.mz.om.oz.dH Ammoo an Asamv m .oz.zs.Hs.qH.zH.Hz.mov emmons loo.>3.os.4d.nz.sz lemme em Asfloav sm .mo.eo.Hm.¢z.e>.mz.mzo some “do 4 Ass 4 «suoaomsoz shadow mmmzzcmzom m0 ZOHBDmHMBMHO AGZOHUWm m osooe CHAPTER VI SPP DEVELOPMENT MODEL: OPERATIONALIZATION AND MEASUREMENT ISSUES The theoretical model is the framework used to empirically examine the influence of solar energy technologies on the development of soft path preferences (SPP). A critical aspect of testing theoretical thought is to translate the abstract concepts into measureable indicators. In this chapter I discuss the indicators chosen to empirically examine the SPP development model. Findings from previous research and the logic used in choosing particular indicators are specified. Issues of index construction, scale development, and other measurement topics are addressed. OPERATIONALIZING THE SPP DEVELOPMENT MODEL In Table 10, the indicators used to operationalize the SPP development model are specified. The items used for the indexes are listed in Appendix B. Table 10 132 133 Table 10 SPP DEVELOPMENT MODEL INDICATORS ENERGY VULNERABILITY NEEDS: (solar - nonsolar samples) 1. 2. 3. 4. 5. 6. 7. Dwelling type (both) - HOUSE Monthly energy costs (both) - AVCOST Number of children in the household (both) - NUMCHIL Household characteristics (both) - FAMTYP Age of respondent (both) - AGE Perceptions of the "national energy crisis" (both) - LADDER Life style impacts of the ”energy crisis" (both) - IMPACT CONTEXTUAL RESOURCES: (solar - nonsolar) 1. 12. 13. Educational level of respondent (both) - EDUC Gross household income (both) - INCOME Personal network support index (both) - PERSUP Institutional support index (both) - INSTSUP Perceived governmental support index (both) - GOVSUP Federal support (nonsolar) (index - solar) - FEDERAL State support (nonsolar) (index - solar) - STATE additionally, for solar only -- Local support index - LOCAL Experiences of personal support index - GRMEMBR Problems of increased cost index - PRCOST Problems of local policies index - PRLCL Problems with people index - PRPEOP Problems acquiring the system - PRBLD TECHNOLOGY CHARACTERISTICS: (solar: active - passive) 1. 2. 3. 4. 5. 6. 7. Years of ownership (both) - OWNLNGTH Problems experienced operating the system (both) - PROPNG Date of system installation (both) - ACTYR, PASSYR Building methods index (both) - ACTBLDMTD, PASBLDMTD Sources of materials and parts (both) - LCLAPRT, LCLPPRT Average number of down periods (both) - ADOWN, PDOWN Cost of solar energy systems (both) - COSTTECH EVALUATION OF SYSTEMS: (solar - nonsolar) 1. 2. 3. 4. 5. 6. 7. General evaluation of systems index (both) - GENERAL Technical capabilities index (both) - TECEVAL Expected energy savings (both) - EXPSAVE Intentions about adding systems (both) — ADDTECH Technical systems support index (both) - TECHSUP also, for solar only (active - passive) -- Average evaluation of systems' condition (both) - ACTCOND PASSCOND Recommended solar to others - RECMEN SOFT PATH PREFERENCES: (solar - nonsolar) 1. 2. 3. 4. 5. 6. 7. Energy self-reliance index (both) - EYSLFREL Local self-reliance index (both) - LCLSLFREL Resources conservation attitudes index (both) - CNRBLF Resources conservation behavior index (both) - CNRBHVR Energy conservation behavior (both) - EYCONS Personal comfort and prestige index (both) - STATUS Personal economic gains index (both) - ECON 134 Energy Vulnerability Needs: Energy vulnerability needs is the driving variable of the model identifying the needs the householders perceive which may lead to a change in the energy base of society. Needs is operationalized in three ways: (1) physical structure of the home, (2) household composition and size, and (3) perceptions of the "energy crisis" and its impacts. One of the important sources of energy needs of a household is determined by the size of its dwelling. Sackett (1984) notes that from 1950 to 1979, the size of the houses in the U.S. increased from 983 square feet to 1,750 square feet. Therefore the heating space actually was higher in 1980 than thirty years earlier. While the sizes of the buildings were growing, there was a decrease in the rate at which energy use grew both in the residential and commercial sectors. Between 1950 and 1973, buildings' energy use grew at about 4.5% per year while between 1973 and 1980, the growth occurred at only .4% a year (Hirst et al., 1981:11). The residential energy use per household declined at an average rate of 1.7% per year between 1973 and 1980. Most of the decline was due to conservation measures such as setting the thermostat lower (40%) and about 20% was due to increased technical efficiencies such as adding attic insulation to homes (Hirst et al., l981:8). 135 While size of dwelling is the ideal indicator, this is only approximated in these data by the type of housing owned by the respondents. Single-family dwellings use more energy than other types of housing, including multifamily dwellings and mobile homes (Morrison and Gladhart, 1976: Newman, 1982). In these data, housing types are: (1) single-family dwelling, (2) multifamily dwelling, (3) mobile home, and (4) other. The average energy cost of the household is computed from average monthly energy bills in winter and in summer. This indicator is only an approximation, for the reported average monthly costs of electricity, heating, hot water and cooling for winter and summer are added together and divided by two. Many solar and nonsolar homeowners verified their costs with bills or with other documents so the responses are fairly reliable approximate measures of actual energy costs (Farhar-Pilgrim and Unseld, 1982:417). Ideally, the energy cost information would be available for the solar homeowners from before the time they acquired the solar technologies. It is a less useful indicator of energy need for the homeowners who already own the solar technologies. As seen previously (Table 3, p.117) the differences in average reported energy costs between the solar and nonsolar 136 homeowners is minimal (e.g. $93.54 to $103.00). Those households living in larger dwellings and paying higher energy costs, will experience greater energy vulnerability needs. The size and social structure of the household also affect the amount of energy used. This is operationalized using number of children, type of household composition, and the age of the respondent as indicators. Several researchers (Morrison and Gladhart, 1976:Newman, 1982) found that larger families and the families in the child-rearing stages use more energy than families without children or at earlier or later family stages. ”Teenage children are the biggest drain on energy resources” (Newman, l982:3). The greater the number of children in the household, the more energy is needed. This indicator and that describing the household composition are proxies for the actual number of people living in a house . . which is the ideal size indicator. The size and estimated energy use of the household are used to rank order the family types. Two parent families are assumed to be the largest, and the multigenerational families the second largest households. Married couples are also great users of energy, especially if both work outside 137 the home. Finally, single parents and single adults use less energy than other household types for they have fewer people. Age has been associated with consumption of energy. Newman (1982) found that age was important because of its association with size of household. There is a sharp increase in energy use in the thirty-five to forty-four age categories. Most energy is expended in the 44 to 59 age categories with a slow decline after this. Energy expenditure and consumption slopes increase steeply in the early phases and decrease less sharply after the 44-59 age period. Therefore the age indicator is ordered based on energy consumption, from the most energy consuming ages to the least as follows: (1) 35-44 years, (2) 45-54, (3) 55- 64, (4) 25-34, (5) 65 and older and (6) 24 or less. Finally, the respondent's perceptions of the national "energy crisis” and of its impacts on the household's lifestyle are used as indicators of perceived national energy vulnerability. National vulnerability perception is computed from the respondents' perceptions of how good or bad the national energy situation would be in 1985 minus the perceptions of what it was like in 1980. The perceptions are based on a ten point scale with 0 - the worst possible situation and 10 being the best. The responses are ranked from perceptions of a worst situation in 1985 than in 1980 138 to a better situation at that time. Finally, the energy vulnerability needs concept also includes an indicator of perceived impacts of the energy crisis on the household's life style. Greater perceived impacts are associated with greater vulnerability needs (Appendix A, Table 25). The solar and nonsolar homeowners' perceptions reflect different behavioral experiences with energy. The solar homeowners have already ”done" something about their perceptions by having acquired solar energy technologies. Obviously it would be preferable to have an indicator directly measuring the perceived need to change the energy base of the household and society, but this is available only indirectly in this data set. Contextual Resources: The second major construct of the model is contextual resources. These resources affect the household's ability to resolve the perceived need for energy by providing barriers or assistance for the resolution of the "needs”. Resources are composed of three parts: (1) the individual's or household's characteristics which serve as resources, (2) resources available in the household's social context, 139 and (3) problems which the solar households have with the solar technologies as described below. The individual and household's resources include: (1) the education of the respondent, and (2) the gross annual household income. The external resources available in the household's social context include: (1) perceived personal network support, (2) perceived institutional support, and (3) types of governmental support. These indicators are shared by both solar and nonsolar homeowners. Both education and income are ordered from highest to lowest values indicating greater to fewer access to resources. Education is measured in this study as the respondents' having completed: (1) graduate studies, (2) college, (3) some college, (4) trade or technical college, (5) high school, and (6) less than high school. Higher education is associated with a higher degree of ability in seeking new knowledge and in using public resources. It has also been associated with a greater awareness of the finiteness of natural resources. Those with higher education (e.g. some college or more) have tended to be more aware of and use federal tax credits for conservation and solar energy technologies (Farhar-Pilgrim and Unseld, 1982: Warkov, 1979: Carpenter and Chester, 1982). 140 However, Warkov (1979) found that although educational level was important in recruiting to the HUD program he examined, it made "little difference as to which households do or do not actually acquire solar" (Warkov, 1979:50). This was similar to Carpenter and Chester's (1982) findings which indicated a strong interaction of income with education. Among homeowners with less than $20,000 income, as education increases so does the percent that filed a tax credit. For homeowners in the $20,000 to $30,000 bracket the effect of education on filing a tax credit is so slight as to be negligible. Among homeowner that have more than $30,000 income, as education increases there is a modest increase in filing for a tax credit. (Carpenter and Chester, 1982:8) Thus the tax credit incentives seem most significant to the middle income persons with higher education and to a lesser extent to those with high income and high education. Education has usually been positively associated with concerns about energy conservation and environmental issues (Olsen, 1983; Farhar-Pilgrim et al., 1979: Farhar et al., 1980; Farhar-Pilgrim and Unseld, 1982; Hamm et al., 1982). Dunlap and Olsen (1984) found that high education was a characteristic of activists of both the hard and soft path ideologies. Many studies have documented the higher educational level of pro-environmentalists (Morrison and Dunlap, 1985: Vine, 1980; Mitchell, 1980; Van Liere and Dunlap, 1980: Farhar-Pilgrim et a1, 1979: Farhar-Pilgrim and Unseld, 1982). 141 Education is also related to concerns about energy. Hamm et al. (1982) found a positive relationship between energy- related attitudes and practies and to beliefs in technological solutions. Farhar-Pilgrim and Unseld (1982) found a positive relationship between education and favorability toward solar energy technologies. However, they also report that those with college degrees and above are more likely to see that solar energy technologies are not a good investment. Farhar et al. (1980: Farhar-Pilgrim et al., 1979) found in their review of the literature that increased education was not associated with favorability toward solar energy technologies. So the role of education as a resource for the use of solar energy technologies is still unclear. Income also is associated with having more resources for acquiring the renewable energy technologies. In this study, household's annual gross income is measured from over $55,000 to under $6,999 a year (Appendix B, p.283). Warkov (1979) found that households which participated in the HUD solar technology programs had higher income than the average household in Connecticut. That pattern was also true of Carpenter and Chester's (1982) findings that those employed part time had the highest percentage of people filing claims for tax credits for solar energy technologies, 'while those who were fully employed were second and the 142 retired, third. Farhar-Pilgrim and Unseld (1982) found a positive association between income and conservation actions, while Hamm et al. (1982) found only a weak relationship between income and energy conservation attitudes and actions. Income has an interesting relationship to environmental concerns. Dunlap and Olsen (1984) clearly indicate that "hard path activists report far higher incomes than do the soft-path activists" (Dunlap and Olsen, 1984:425). Morrison and Dunlap (1985) examined this issue further finding that the modal income of members of environmental organizations was above average. Van Liere and Dunlap (1980) report that income is often negligible or sometimes negatively related to environmental concerns. They also report that favorability to solar energy technologies increases with income, including the probability of investing in a solar technology (Farhar- Pilgrim and Unseld, 1982; Farhar-Pilgrim et al., 1979: Farhar et al., 1980). The majority of the findings suggests that income is positively associated with favorability toward and use of solar energy technologies. Therefore, I expect income to have a positive relationship to the contextual resources 143 variable, to use of solar energy technologies and to the development of soft path preferences. The contextual resources for the solar energy technologies variable also indicates some of the conflicts occurring when households change energy base. This variable is the primary one with indicators to capture some of the societal conflicts suggested by soft path critics. Personal network support is indicated through the respondent's perceptions that friends, neighbors, job associates, and spouse or housemates are in favor of solar energy technologies (Appendix B,p.283). This is an additive 17 point index which is weighed making the spouse or housemate's answer most important, neighbors and friends intermediate, and that of the colleagues less important - see Farhar-Pilgrim and Unseld (1982:179). Technology support institutions may directly assist or create barriers to the acquisition of solar energy technologies. The additive 20 point index of institutional support provides a measure of how well institutions support alternative energy systems: (1) utility companies, (2) homeowners' insurance companies, (3) technology sellers through warranty coverage of systems, and (4) solar energy companies' dependability (Appendix B,p.284). 144 The third contextual resource examined is governmental support. The nonsolar and solar respondents were asked whether or not they perceived the three levels of government (local, state and federal), as "favorable”, "mixed” or "opposed" to the use of solar energy technologies. The responses were weighed providing stronger influence to local governmental support, reflecting Lovins' emphasis on community based action. The additive 18 point index of general governmental support is another indicator of contex- tual resources support for solar systems (Appendix B,p.284). The nonsolar homeowners reported whether or not they knew if there was a federal or state income tax credit for the purchase of solar water and home heating technologies. The solar homeowners indicated their participation in federal demonstration programs, and whether or not they received local and state property and sales tax credits or exemptions (Appendix B, p.285). The number of grants, credits, or exemptions received is summed creating a four point index of state and federal support, but only a three item index was possible for local support. Carpenter and Chester (1982:5) found that the most important variable in determining whether homeowners had solar water or space heating technologies were tax credits. Respondents claimed that these technologies would not have been installed without the tax credits. 145 Additional indicators are available for the solar sample. A three point personal membership in solar voluntary associations or professions is used (Appendix B, p.288). It is assumed that people who participate in groups focused on solar energy technologies are linked into a network providing more resources for changing to renewable energy technologies, so this is expected to provide additional resources for the use of solar energy technologies and the development of soft path preferences. Actual problems experienced with the solar technologies are indicators of contextual barriers for the solar homeowners. Problems in choosing between solar systems, lack of information, difficulty in choosing contractors, builders, or architects, increased costs, policy problems and people problems and installation issues are specified. A five point index is also created for increased costs due to difficulties in obtaining financing, increased property taxes, and utility rates, and decreased resale value of the home (Appendix B, p.286). When homeowners experience negative economic factors because institutions in their communities increase charges, they will perceive greater barriers and fewer resources available for ownership of alternative energy technologies. Problems with solar access rights and local building codes create an additive three point scale indicating local policy 146 barriers. A four point additive scale of people problems is created from reported experiences with vandalism, neighbors' opposition, and obstruction of solar technologies. The greater the policy and neighborhood problems experienced, the fewer the perceived resources in the community. Finally, an additive five point index is created from problems experienced in attempting to acquire alternative energy systems (Appendix B, p.288). The contextual resources variable therefore is composed of indicators which both facilitate and detract from the ability and favorability of the householders to acquire solar energy technologies and develop soft path preferences. The homeowners may find that the resources available in the community are not sufficient to adequately support their need for renewable energy systems. The more problems they encounter, the fewer the resources in their locale. The fewer the problems, the greater the resources. If the resources are perceived as being adequate for the resolution of the perceived energy vulnerability needs, then the homeowners may acquire solar energy systems. The use of solar energy technologies is expected to be positively associated with the development of soft path preferences. 147 Technology Type: Technology type is a very important variable in the development of SEP according to Lovins. Information is provided only by solar homeowners for this concept. The years of solar technology ownership is computed from the number of months of ownership divided by 12. The date of acquisition is also reported separately for the active and passive systems (Appendix B, p.289). The date the technology was installed or constructed reflects different cohorts' experiences with social contextual influences on the development of SPP and with the operation of the solar technologies. The length of ownership indicates how long energy has been visible to the owner or that the homeowner has been associated with a solar network. Due to cognitive dissonance, she/he may become more committed to the values and attitudes of the soft path changes reflected in soft path preferences. Thus, the longer the ownership length, the greater the commitment to soft path preferences. Length of ownership also is associated with the experiences of the homeowners with the operation of the solar energy technologies. The longer the ownership length, the greater the possibility for having problems operating the solar technologies. The problems experienced operating the solar energy technologies form a five point summated index measured with indicators of: (1) maintenance problems, (2) difficulty finding paths or equipment for repair, 148 (3) problems with overheating, and (4) problems from poor system design (Appendix B, p.289). As indicated previously, Lovins (1976) claims that those who obtain solar technologies are interested in taking control of their own energy systems. Therefore important characteristics of the solar technologies include whether the owners make them from local resources. A ten point. additive index was developed from indicators about how the energy systems were manufactured or designed, built and installed. The indicators are weighed according to whether or not the technology was created by the respondent locally or by a foreign firm or hired professional (Appendix B,p.290). Because of soft path's emphasis on self-reliance and control of one's own energy systems, the following characteristics of the manufacturing or building process are more consistent with the soft path: (1) being manufactured by small local firms instead of national and foreign firms; (2) being built on site rather than as complete packaged systems and then assembled, and (3) being installed by the owner as opposed to a hired contractor (Appendix B, p.291). These characteristics reflect how solar energy systems can be installed using processes congruent with either soft path or hard path orientations. The sources of the parts and materials for the solar systems 149 also form part of the technologies' implementation process. The following is the ranking and rationale given for the ordering of an eight point variable: (1) local solar companies are specifically committed to renewable energy systems, so they are identified as being most consistent with the development of the soft path changes: (2) local hardware, lumber or supply companies are general purpose stores supporting "do-it-yourselfers” and technical self- reliance: (3) local heating, plumbing stores are more dependent on items manufactured elsewhere, so they are a little less consistent with soft path developments: (4) the use of local building contractors and architects sources is another step away from direct control and participation by the owner of the solar technology; (5) local outlets of chain stores, mail ordering from nonlocal solar companies and getting materials and parts from non-local solar companies are all steps away from local self-reliance. These indicators of the acquisition process are reported separate- ly for active and passive systems (Appendix B, p.291). The total installed cost of the solar energy technologies is also an important characteristic of the solar technologies (Appendix B, p.289). The technology type as measured by the implementation process, the length of ownership, operating problems and cost of the technologies will affect the evaluation which the owner makes of the alternative energy 150 systems and will influence SPP development. Evaluation of Solar Systems: Evaluation is the third construct in the nonsolar model and fourth for the solar model. It is a combination of general and specific evaluation indicators, and measures of the intentions of the homeowners about future use of solar energy technologies. These questions were asked retrospectively of the solar homeowners, and hypothetically of the nonsolar sample. A ten point additive index reflecting general evaluation of solar systems is created from overall reactions and specific experiences as solar system owners. The more positive the homeowners' assessment, the more inclined they will be to continue using renewable energy technologies and the greater their soft path preferences. A thirty five point additive scale is developed from items about the technical capacities of solar energy technologies in the following categories: (1) technical and economic performance, (2) climatic adaptation to sunny and cold climates, (3) longevity, (4) technical obsolescence, (5) safety of the system and (6) operating reliability (Appendix B, p.292). 151 An assessment of solar energy technologies as a good investment is computed as the difference between expected and actual savings from the solar system. A higher percentage being saved compared to what was expected is associated with a more positive evaluation of renewable technology ownership supporting longer use and stronger SPP development (Appendix B, p.293). The intentions of the respondent were directly addressed with the question "To what extent, if any, have you considered investing in solar energy technologies in the next 2-3 years?” or " adding more solar." Four values were given as possible responses from "will definitely invest" to "have not considered." Solar homeowners have an additional four point indicator of whether or not they have recommended solar technologies to others. The nonsolar homeowners' indicator describes their perceived ease of using solar technologies on their home (Appendix B, p.295). Finally, only solar homeowners are asked to evaluate the: (1) helpfulness of maintenance and operating instructions for the solar system, (2) warranty coverage, and (3) quality of repair service. This creates a 15 point additive index from the Likert-like five point satisfaction scale for each item (Appendix B, p.296). 152 Separate evaluations were made by the solar homeowners of the overall conditions of their active and passive solar systems. These indicators were computed as an average of a five point assessment scale from "excellent" condition with no problems to ”poor systems” with chronic problems (Appendix B, p.297). A positive evaluation of the solar energy technologies is associated with greater soft path preferences development. Soft Path Preferences: Soft Path Preferences is composed of five basic dimensions: (1) self reliance preferences, (2) natural resources conservation preferences, (3) personal gains preferences, (4) equity and (5) social diversity preferences. The last two are not operationalized in this data set and are social impacts of a changed energy base. Self reliance is central to the Lovins' theory of social change. It is through self-reliant attitudes and behaviors that households and communities are able to become independent from the centralized dominant social structures of the national and international levels which control the hard energy path. It is through the process of self- reliance that individuals working in communities can recapture some of the energy resources which are being siphoned away. 153 In these data, self reliance preferences is measured in two ways: (a) attitudes about energy self-reliance, and about (b) local community and personal self-reliance. An additive 20 point index was created from beliefs in the importance of:(1) increasing overall self-reliance, (2) having a more reliable energy supply, (3) reducing the need for large power plants, and (4) increasing independence from utility companies (Appendix B, p.297). A twenty point additive local self-reliance index was created from variables measuring attitudes about whether local communities or the federal government would be able to resolve the energy crisis, about the importance of being independent from national policies, and about whether or not the respondents were developing self-reliance skills (Appendix B, p.298). The second dimension of soft path preferences, natural resources conservation preferences, is important to Lovins' ideas of equity. It is only as energy and other natural resources are conserved by the present generation that there will be some left for future generations and other nations. Energy conservation also provides additional resources for the transition from the hard energy path to the soft energy path. 154 Natural resources conservation preferences uses measures of: (1) general attitudes towards natural resources conservation, and (2) perceptions about the importance of easing energy shortages, and (3) of decreasing personal use of energy resources. These items form a 15 point additive index of the perceived importance of conserving natural resources (Appendix B, p.299). Natural resources conservation behavior will lead the transition to the soft path independently of belief systems, claims Lovins (1976). The behaviors release additional natural resources, not the attitudes. Therefore, a 15 point additive natural resources conservation behavior index is created composed of the reported activities of: (1) recycling household materials, (2) recycling clothes, and (3) contributing to ecological organizations (Appendix B, p.300). Another index of natural resource conservation behavior is created by the number of energy conservation behaviors out of 15 performed by the respondents (Appendix B, p.300). The greater number of such activities is positively associated with SPP development. A strong dimension of the Lovins' argument is that the soft path will be developed as people discover that it is to their personal economic and noneconomic advantage to change 155 their energy base. Economically, solar energy is cheaper than nonrenewable energy, except for the fact that government subsidies make the nonrenewable energy appear to be cheaper. The noneconomic gains include the satisfaction of self-reliance and the attendant self-esteem. In these data, the dimension of personal gain is measured by the importance of: (1) increased comfort and prestige, and (2) increased personal economic gains. Increased personal prestige, and home comfort are measured with a ten point additive index (Appendix B, p.301). Both the hard path and the soft path advocates claim that personal economic self interest is served by ”their" energy base. Personal economic gains are measured on an additive 20 point index of perceptions that: (a) solar is a good investment, (b) it saves money, (c) it is a protection against rising energy costs, (d) it decreases utility bills, and (e) increases the resale value of the home. Several of the dimensions of soft path preferences are inadequately measured in these data. The dimension of self- reliance preference could have more directly measured the attitudes about national energy self-reliance. Community energy self-reliance as opposed to dependence on the federal government was adequately measured with three indicators. 156 Personal and household self-reliance aspects were inadequately tapped with only one indicator. The second dimension of natural resources conservation only had one indicator which did not focus on energy conservation attitudes. A broader use of indicators of other aspects of natural resources conservation, such as water conservation, air quality, throwaway packaging, etc., would have strengthened the attitudinal measure. The behavioral indicators include behaviors which are upper middle class oriented (e.g., contributing to ecologically oriented organizations) as well as lower class oriented (e.g., buying clothing at a garage sale or at a second-hand store). Other behaviors could have been tapped such as the sharing of neighborhood tools, garden composting, reusing aluminum foil, using the bicycle or public transporation, etc. Finally, the energy conservation indicators inadequately measure the bahaviors of serious energy conservers. For example, walking to work, living downtown to prevent the urban sprawl, 223 using appliances or fireplaces, and so on are not measured. The third dimension, of noneconomic personal gain, was measured with only two indicators. One indicator mixed self- 157 esteem with increased status and prestige. Indicators of satisfaction with increased diversity, equity, self-reliance and local control would have strengthened the variable. Measures of the respondents' perceptions of personal and household gains of the soft path, such as increased neighborhood cohesion, would have been useful, also. While offering these criticisms and suggestions for the development of stronger indicators for measuring SPP, I acknowledge that this is not an uncommon problem with secondary analysis. These data were gathered for a diffusion and marketing study of solar energy technologies. They were quite appropriate for that purpose in 1980. MEASUREMENT ISSUES In preparing the data for analysis, the attributes and values of each of the variables were ranked in relation to theoretical and empirical knowledge about the relationship of the variables to the soft path preferences development model's concepts. The lowest rank was assigned to those attributes and values which were most closely related to the development of soft path preferences. In most cases the variables have at least four attributes or values although in a couple of cases it was necessary to retain only three attributes/values. Dichotomous variables 158 were used to develop indexes. Variables which were measured on scales quite different from the 3-50 point range used most frequently, were transformed into scales more compatible with the other variables by dividing through by 10, usually. Indexes were created with as much similarity as possible between the solar and nonsolar samples. The items within the indexes were chosen primarily through face validity though that was empirically verified by examining their inter-correlations and Cronbach split-half reliability scores (Cronbach, 1951). The variables which displayed no correlation with the other variables were deleted from their indexes. The construction of the indexes is specified in Appendix B and in Table 11. Table 11 The primary difference between the solar and nonsolar indexes is that the solar homeowners specified actual experiences with the technologies while the nonsolar homeowners had to respond more abstractly. The table indicates the correlation between the items and their indexes without the item. 159 Table 11 INDEXES OF THE SPP DEVELOPMENT MODEL Correlation Alpha w/o Item to Index Item Scale 8* N* S N Variance RESOURCES: 1. Personal Network support Spouse .153 .397 .482 .387 Friends & neighbors (x2) .469 .487 .133 .301 S: .007 Friends (X2.5) (CRFRNDS) .016 .006 .575 .707 N: .001 Colleagues (x3) .462 .470 .150 .336 alpha S: .575 N: .707 2. Institutional support Utility company .606 .298 .814 .704 Warranty coverage .708 .524 .769 .506 S. .133 Insurance coverage .698 .518 .774 .479 N: .730 Firm's dependability .644 .409 .798 .563 alpha S: .833 : .627 3. General governmental support Federal support (x3) .429 .390 .350 .224 State support (X2) .488 .444 .319 .179 S: .006 Local support .416 .358 .369 .246 N: .594 Local laws (X1.5) (CRLAWS).052 016 .727 .800 alpha S: .727 N: .800 4. Federal support received Federal tax crdit .223 ---- .027 ---- HUD - 1 .003 ---- .364 ---- S: .010 HUD - 2 .212 ---- .022 ---- alpha S: .263 5. State support received Sales tax exemption .076 ---- .020 ---- Property tax exemption .136 ---- .004 ---- S: .011 Low income loan .089 ---- .051 ---- Income tax credit(TAXSTA)-.023 ---- .359 ---- alpha S: .359 8* - solar, N* - nonsolar 160 Table 11 (cont.) Correlation Alpha w/o Item to Index Item Scale S N S N Variance 6. Local support received Sales tax exemption .193 ---- ---- ---- Property tax exemption .193 ---- ---- ---- S:.002 alpha : .239 7. Personal support Member solar groups .352 ---- ---- ---- Solar professional .423 ---- ---- ---- S:.003 alpha : .756 PROBLEMS: 1. Increased cost Financing difficulties .265 ---- .405 ---- Utility rates up .320 ---- .348 ---- S:.0004 Property tax up .257 ---- .423 ---- House resale values down .322 ---- .410 ---- alpha S: .466 2. Local policy problems Solar access rights .226 ---- ---- ---- Local building codes .226 ---- ---- --—- S:.001 alpha S: .266 3. Problems from people Vandalism .264 ---- .215 ---- Neighbors' opposition .241 ---- .234 ---- S:.0007 Obstruction of system .201 ---- .438 ---- alpha : .348 4. Acquisition problems Installation .252 ---- .473 ---- Choosing a system .261 ---- .463 ---- S:.0007 Inadequate information .342 ---- .384 ---- Inadequate professionals .344 ---- .401 ---- alpha S: .502 161 Table 11 (cont.) Correlation Alpha w/o Item to Index Item Scale P** A** P A Variance TECHNOLOGY TYPE (solar) 1. Building methods used Manufacturer/designer 1 .852 .576 .929 .826 Manufacturer/designer 2 .981 .707 .902 .872 A:21.322 Construction method 1 .845 .708 .941 .809 P:15.909 Construction method 2 .975 .852 .893 .754 Installation method 1 .869 .591 .926 .818 Installation method 2 .987 .851 .891 .755 alpha A: .831 P: .930 2. Problems operating the systems (solar) Maintenance problems .308 ---- .315 ---- Repair parts .246 ---- .382 ---- S:.0008 Overheating .126 ---- .479 ---- Improper system design .333 ---- .292 ---- alpha S: .446 EVALUATION: 1. General evaluation Favorability to solar .245 .574 ---- ---- S:.041 Satisfaction with solar .245 .574 ---- ---- N: .101 * alpha : .389 N: .718 2. Technical capabilities Economic and technical .251 .225 .618 .559 Warm climates only .240 .211 .623 .565 S:.226 All climates good .309 .359 .604 .509 N:.154 Longevity .316 .258 .605 .549 Obsolescence .402 .306 .571 .531 Safety .406 .338 .569 .518 alpha : .629 : .573 3. Technical support Helpfulness instructions .092 ---- -.244 ---- Warranty satisfaction .084 ---- -.255 ---- S:3.275 Repair satisfaction .155 ---- .569 ---- alpha S:-.037 P** - passive, A** - active 162 Table 11 (cont.) Correlation Alpha w/o Item to Index Item Scale S N S N Variance SOFT PATH PREFERENCES 1. Energy self-reliance Increased overall .636 .629 .661 .705 Var. Reliable energy supply .575 .640 .688 .705 S: .063 Independence of utilities .573 .564 .689 .741 N: .044 Fewer large power plants .446 .526 ..757 .758 alpha S: .756 N: .780 2. Local self-reliance Use community resources .095 .067 .286 .238 Var. Independence of federal .203 .131 .150 .166 S: .242 Trust federal policies .148 .124 .232 .176 N: .535 Self-reliance skills .141 .146 .241 .144 alpha S: .289 : .235 3. Resource conservation beliefs Ease energy shortage .655 .465 .440 .328 Var. Conserve natural resources.642 .469 .463 .331 S: .313 Reduce consumption .356 .231 .799 .665 N: .041 alpha S: .709 : .572 4. Resource conservation behavior Recycling paper, glass. .250 .202 .208 .144 Var. Contribute to ecological .268 .178 .172 .192 S: .681 organizations N: .484 Buy second hand goods .164 .136 .386 .277 alpha S: .371 : .295 5. Personal comfort and prestige Greater home comfort .167 .206 ---- ---- Var. Increased status and S: .378 prestige .167 .206 ---- ---- N: .055 alpha S: .277 N: .340 6. Personal economic gains Protection from increased energy costs .511 .563 .619 .711 Var. Reduced energy bills .608 .625 .544 .678 S: .102 Long-term savings .623 .668 .544 .661 N: .105 Increased home resale .254 .458 .791 .791 alpha S: .695 : .764 163 Because of the low correlation with their indexes, and because of the way they depressed the reliability of the index, some of the items were dropped. This was true for the "personal network support index" (dropped "concerned about what friends and neighbors would say”): for the ”general governmental support index" (dropped "concerned that codes or covenants might prohibit use of solar energy technologies in your home"), and ”state support received index" (dropped "received state tax credit"). Items were deleted only if they were weak in pggg the solar and nonsolar samples and if they were nOt central to the construct. Those items which were theoretically central for construct validity were retained even though this reduced the reliability of the index as a whole. See, for example, the index "technical support“. In developing the indexes and preparing the data for modeling, the missing values were included as such if they were under 5% of the answers. If 5-20% of the answers were missing, the mean value of the variable was substituted for the missing values. If the indicator had a higher percentage of missing values, it was deleted from the analysis. Pairwise deletion of missing values was used in the principal components and Pearson correlation analyses. 164 The "don't knows" were sometimes coded as a "neutral” value in the ranking of the attributes or as ”missing values" depending on the particular variable. This can be examined in the frequency tables of Appendix A. Sometimes the frequencies were highly skewed because of a few outliers. The distributions were capped at three standard deviations from the mean if there were few respondents in those categories. These can also be examined in Appendix A. In order to compare the degree of soft path preferences which developed among the various samples -- solar, nonsolar homeowners: passive - active solar owners: those who had owned solar technologies for different periods of time -- a soft path preference scale was created. The variables of the soft path preferences were factor analyzed using principal components extraction with orthogonal varimax rotations. The communality values were used as weights in developing scores for each respondent. Each indicator of the soft path preference was multiplied by the weight of the variable and then standardized by using the sum of weights as the denominator. 165 Table 12 SPP SCALE WEIGHTS: SOLAR AND NONSOLAR Variable: Solar Nonsolar Energy self reliance .77772 .81407 Local self reliance .61171 .44332 Resource conservation beliefs .77289 .63946 Resource conservation behavior .65177 .43049 Energy conservation behavior .35973 .55338 Personal comfort and prestige .62019 .53421 Personal economic gains .74135 .75970 In order to prepare the SPP development model, preliminary examination of the data was performed by factor analyzing the theoretically defined indicators of the energy vulnerability needs and the contextual resources concepts to determine if the variables were empirically associated with those two concepts. The same process was followed with the concepts of technology type, evaluation, and soft path preferences. The results of the orthogonal and Oblique rotations were identical in loading the variables on similar factors with very similar weights. This clearly identified the factors adding confidence to the use of the partial least squares modeling procedure which uses principal components extraction. 166 As a result of the analysis, "respondent's age" was found to be more closely associated with variables of the energy vulnerability needs concept than with the contextual resources variables where it had been originally placed. It therefore became an indicator associated with energy vulnerability needs. The ”behavioral intention" variables were strongly associated with the evaluation variables, so they were also used as indicators of this construct as argued in the theoretical section. "Problems of increased cost" and the "real savings" from solar technology use were more closely associated with evaluation concept variables than with soft path preferences variables, so they were moved to that concept. The following variables were dropped from further analysis because they did not correlate very strongly with other variables. "Moving plans” and "type of primary heating fuel” were both dropped from the energy vulnerability needs concept. Other variables which were dropped due to large missing values or lack of clear construct validity were: (1) "types of solar technologies known to the respondents" (Farhar-Pilgrim and Unseld, 1982), (2) "willingness to take out a loan to pay for solar technologies”, and (3) ”participation in local community activities”. Finally, results from Pearson correlation analysis were used to enter the variables into the partial least squares (PLS) 167 modeling program (Appendix C, Tables 38 and 39). CHAPTER VII ANALYSIS In this chapter, I describe the partial least squares (PLS) covariance structure model which I used as a method for examining the interrelationships of the indicators with the concepts of the SPP development model and to describe the relationships among the model's concepts. These relationships will be described for four major subgroups. The solar and nonsolar samples will be compared; then the active and passive solar system owners will be compared. This will help determine if the process of soft path development is similar or different for the four groups. After the theoretical model has been examined using indicators which are as similar as possible for each subset being compared, then the model is trimmed to express the relationships more parsimoniously. In order to facilitate the discussion of the method of analysis, I will use the terms used most often by structural equation modelers. PLS is one of that class of methods. The theoretical concepts which form the primary variables of the model are called "latent variables" because they are not measured directly. The measureable items which I have called indicators are labelled "manifest variables" by structural 168 169 equation modelers. INTRODUCTION TO PLS Partial least squares structural equation modeling was developed by Herman Wold (1980) in response to some of the situations researchers find themseleves in - searching to analyze relationships between variables when theoretical knowledge is scarce and when the researcher is not sure of the theoretical distribution of the population being studied. This method was developed as an alternative to the "hard modeling" approaches (Falk and Stuber, 1984) which are primarily aimed at accuracy and model testing. PLS soft modeling is "intermediate between data analysis and the (hard) assumptions of the ML [sic. maximum likelihood] mainstream of contemporary statistics" (Wold, l982:5). PLS is designed as a complimentary tool which maximizes consistency, or patterns, rather than accuracy (Wold, 1982:53). The model standardizes the latent variables to unit variance. Using least squares, PLS also minimizes all residual variances jointly without optimizing total residual variance or using other criterion of optimization (Wold, 1980:67). This is a major difference between partial least squares operations and maximum likelihood operations whereby all of the residuals are minimized simultaneously. PLS does it iteratively by block of manifest variables (Lohmoller, 1984). 170 The PLS method also does not face identification problems because it explicitly estimates the latent variables from the weighted aggregates of their manifest variables with weights determined by the weight relations mode (Wold, 1980, 1982; Lohmoller, 1984). The manifest variables are therefore quite critical to determining characteristics of the latent variables. In PLS, the noise from random variability is reduced when the loadings Of the manifest variables are large. The residuals from one latent variable ”. . . are the data input for PLS estimation . . .” of the next latent variable as the program goes through the various iterations (Apel and Wold, 1982:222). Apel and Wold explain: Designed primarily for research contexts that are simultaneously data rich and theory-primitive, soft modeling has to cope with noise that is inevitable in the indirect observation of latent variables by manifest indicators, plenty of noises when only some few indicators are available, each of which carries noise of its own. (Apel and Wold, 1982:237). Apel and Wold (1982) recommend using loadings of between 0.5 and 0.6 though they also use 0.3 (Apel and Wold, 1982:229). There are three weight relation modes which determine the manifest variables' relations to the latent variables. Through Mode A, the relationship arrows connecting the 171 manifest and latent variables are pointed outwardly from the latent variable. This mode is consistent with principal component analysis for the first latent variable. Although the PLS estimates are more stable with Mode A, which is basically a sequence of simple OLS regressions (Apel and Wold, 1982:228), this mode tends to reduce the beta values between the latent variables by creating minimum residual variances in the block structures of the latent variables (Wold, 1980:70). Using Mode A, the PLS procedure minimizes, in its first stage, each residual variance to estimate the loading on each manifest variables one by one (Wold, 1980:67). This can be interpreted as the contribution of each manifest variable separately to the latent variable. Mode B is the second mode and is less stable for it is more like multiple regression or canonical correlation for the first latent variable or canonical variate (Levine, 1977). Thus it provides the best possible prediction of the latent variable without regard for residual variance on the manifest variables (Falk, 1986). Mode B is usually indicated by arrows coming into the latent variable from the manifest variables. Therefore, the manifest variables' weights are the amount which each variable contributes jointly with the other manifest variables to the latent 172 variable. This mode tends to increase the beta values and the correlation between the latent variables that are connected with the arrow scheme (Wold, 1980:71). Wold (1980) recommends using Mode B for exogenous latent variables, but using Mode A for the other latent variables. Bielby and Hauser (1977) suggest that the Mode A should be used when the manifest variables "appear as indicators (reflections or effects) of latent variables . . ." while the Mode B is used when ”. . . observables appear as causes of latent variables" (Bielby and Hauser, 1977:147). Mode C is a combination of A and B. In the PLS program (Lohmoller, 1984) only one mode can be specified for each block of variables. One of the advantages of using PLS modeling is that it can be used to analyze: (l) dichotomous variables, (2) categorical variables, (3) aggregated categorical variables, and (4) internal or ratio variables. Since these can be included together in an analysis (Lohmoller, 1984; Falk, 1986), there are no restrictive assumptions about the measurement level of the data. PLS assumes linear relationships between the manifest and latent variables, and between latent variables. It also assumes that the manifest variables' residuals are 173 independent of each other and of the residuals of the latent variables. An advantage of the PLS program is that it actually examines the correlations between the residuals and the various other parts of the model. The data used in this analysis meet the assumptions of the modeling program. A criticism often made of PLS is that it provides a biased estimate of the relationships of the model. In fact Apel and Wold (1982) developed an experiment indicating that PLS estimates are indeed biased . . . "the tendency being on the one hand to underestimate the LV [sic. latent variable] correlations and path coefficients" while overestimating the relationships between the manifest variables and the latent variables (Apel and Wold, 1982:223; Dijkstra, 1983). Areskoug (1982) examined the impact of the bias and concluded that "the loadings preserve their relative magnitudes within blocks, and the estimated correlation between the latent variables is the highest possible among linear forms of indicators" (Areskoug, 1982:106). The manifest variables are fed into a PLS computer program in block order, sequentially reflecting the model. These blocks of manifest variables create the latent variables and are the theoretical relationships of the model. The direction of the relationships between the latent variables 174 is also specified. The influence of the manifest variables operates solely through the latent variables (Lohmoller, 1984; Wold, 1982). ComparingPLS and ML: Areskoug compared the performance of PLS and maximum likelihood estimation of factor models (ML) which is the dominant mode in structural equation modeling. One of the disadvantages of ML is that it is based on heavy iterative procedures and thus may be very expensive for large samples. ML is " . . . especially designed for situations with detailed knowledge of the data and the model structure" found in experimental fields (Areskoug, 1982:107). PLS, however, "attempts to derive robust estimates under a minimum of assumptions about the stochastic properties of the model . . ." by focusing on the relations between the latent variables (Areskoug, 1982:96). Bentler and Weeks (1980) claim that the greatest advantage which PLS has over ML is that estimates can be obtained where they cannot with ML because of the assumptions of normality and of the costly iterative computer processes of ML. Joreskog and Wold (1982) compared the two analytic processes, concluding that they are complimentary. "ML is 175 theory-oriented, and emphasizes the transition from exploratory to confirmatory analysis. PLS is primarily intended for causal-predictive analysis in situation, of high complexity but low theoretical information" (Joreskog and Wold, 1982:270). Through a dialogue with the computer, PLS helps consolidate, improve and further develop the model. The focus in ML estimation is simply on whether or not the model fits the data. Rationale for Selecting PLS Rather than ML Modeling: One of the assumptions which "hard modeling" makes is that one has a closed theoretical system with all of the concepts correctly specified and with no important ones omitted (Falk and Stuber, 1984: Falk, 1986). As indicated previously, the model which I have specified about soft path preference development does omit the important theoretical variable of social structural changes and its feedback to soft path preferences, perceptions of energy vulnerability needs and contextual resources. Clearly I am not meeting the first criterion of hard modeling. Furthermore, even within the portion of the theoretical model which I am proposing to examine, there are ommissions - specifically, equity and social diversity values and attitudes were left out due to data limitations. 176 Other limitations include the poor construct validity of some manifest variables, such as using number of children and type of family for size of the household. Other problems with the manifest variables have been discussed in the operationalization and measurement chapter. As mentioned in the operationalization chapter, some of the additive indexes have poor split-half reliability scores. Although this is not fatal in additive indexes, it does present another measure of how well the individual items making up the indexes are related (Cronbach, 1951). They do not meet the criterion of being well tested empirical measures of theoretical concepts. Also, the data are primarily ordinal and not interval as required by maximum likelihood estimators. Furthermore, the solar sample was a snowball sample. Although it is very large, it was not randomly selected and many of the variables are not normally distributed as required in "hard modeling". Although large samples provide many advantages, they are also very expensive when all manifest and latent coefficients are estimated simultaneously (ML), particularly if the model has any complexity. Although accuracy is a strong contribution of maximum likelihood estimations, this is not the primary purpose of 177 my current research. Since the manifest variables are only approximate measures, it seems foolish to provide a highly accurate estimate of their relationships! I am more interested in examining the comparative strengths of the different relationships in the model to discover if the strength of those relationships is consistent with what is predicted by the theory of soft path development. I have proposed a model of theoretical relationships which combines well developed research traditions with one having little empirical research. There are still many questions about how the manifest variables are related to the latent variables of the soft energy path. Thus the proposed system of relationships needs to be explored and refined. PLS allows exploration of nested models using chi-square for comparing the relative fit of different models rather than as a probability test of "goodness of fit". This is especially true with very large samples (Bentler and Bonett, 1980; Falk, 1986). This exploration is conducted conservatively when compared to ML estimations for PLS tends to overestimate the measurement model's (e.g. manifest to latent variables) relationships and to underestimate the relationships among the latent variables. 178 A These are the primary reasons why I have chosen to use the partial least squares process for examining the theoretical model. Although ML is ideal for comparing well developed models since its mathematical properties are better understood and it is more precise, I do not believe it is appropriate for this project. PLS SPECIFICATIONS OF THE SPP DEVELOPMENT MODEL The model describing the theoretical process of soft path preferences development was examined by comparing the solar and nonsolar samples using similar manifest variables. Then the model was respecified for the solar homeowners to include the additional latent variable "technology type" and some manifest variables specific to the experiences of the solar system owners. The second model comparison was made between two subsets of solar owners, the active and the passive solar system owners. Equivalent manifest variables specified to their particular types of solar systems were used in the modeling process. In every comparison, a measurement model was first run. This specified the number of latent variables in the model, but put all variables into one block so that the manifest variables were free to associate with any latent variables. 179 Measurement Model for Solar and Nonsolar Homeowners: Originally, 29 manifest variables were entered into the measurement model for the nonsolar homeowners and 43 into the solar homeowners model in the sequence specified in Tables 17 and 18. In order to keep the models comparable, the only manifest variables which were deleted were weak for both samples. The very conservative criteria used were: (1) Were the loadings (Mode A) of the manifest variable under .300? (2) Did the manifest variable have a low score on both the weights and the loadings (Mode B and A)? (3) How theoretically central was the manifest variable to the latent variable? There were two manifest variables which fit all the criteria and therefore were deleted from further consideration: (1) "perceived governmental support" (solar-.214; nonsolar=.164) and (2) "institutional support" (solar-.048 and nonsolar=.098). Two additional manifest variables were deleted from each of the separate sample's models. For the nonsolar sample, the "estimated acquisition and installation cost of a solar water heater" (.229) and the estimate for a solar home heating system (.281) were deleted because they had very high missing values. 180 Although the manifest variables indicating the methods used in building or acquiring the active or passive solar systems had low values (.178:.226), their theoretical centrality served to retain them in the solar models. For the solar sample, "expected payback" (-.009) was deleted from the model due to its low association with the latent variable. The "expected payback" variable was computed from the respondents' estimates of the cost of solar systems divided by their average yearly energy costs. It simply was not a sufficiently strong indicator being primarily composed of "noise". After these deletions, the nonsolar sample's model was composed of 25 manifest variables and the solar sample had 40 manifest variables. These formed the foundation for exploring the process of SPP development. Measurement Model for Active and Passive Technology Owners: Forty two manifest variables were placed in the active and passive measurement models. Several were removed because they had extremely high missing values and were therefore not considered further. The first indicating the "source of parts for passive and active systems" (missing data - 47%), the "number of times the passive solar systems had been unoperative" (missing data I 99%), and finally "how long the 181 respondents felt the solar systems would last” (missing data -87%). "Institutional support" and "perceived governmental support" were dropped for consistency with the solar and nonsolar samples. As a result of the measurement model run, two more manifest variables were dropped - "personal network support" (solar = -.041: nonsolar --.099) and "problems operating the system" (solar --.020: nonsolar --.047). Therefore, the passive and active solar SPP development models were completed with 35 manifest variables. Specification of the Theoretical Model: An important influence on the path relationships between the latent variables is the mode used in relating the manifest variables to them. The mode specifications of the manifest variables are indicated in Table 13. Although it is possible to use one manifest variable per latent variable, it is better to use at least three (Falk and Stuber, 1984; Falk, 1986; Kim and Mueller, 1978). This criterion has been met. Table 13 182 Table 13 MODE SPECIFICATIONS FOR THE SPP DEVELOPMENT MODEL Latent Minimum no. Variable Mode* Indicators Energy vulnerability needs B 7 Contextual resources B 5 Technology type (solar) A 4 Problems (solar) A 5 Evaluation of systems A 6 Soft path preferences A 7 Model type C 25** *Mode: A - outward directed B - inward directed C - mixed mode **The minimum number of indicators is associated with the nonsolar model. Since the smallest sample size used in the analysis is that of the passive solar technology owners (n-326) this is still well within the criterion of 5 respondents per manifest variable. The largest number of variables analyzed using the passive subgroup is 37. The soft path preferences development model allows me to examine the influence which the manifest and latent variables have on the development of attitudes and behaviors consistent with social structural changes labelled the soft energy path. There are three measures which summarize how completely the 183 model captures the data:(1) chi-square without taking into consideration the effects of the model: chi-square with the model, (2) the multiple correlation (R-square) for each endogenous latent variable, and (3) the Bentler - Bonett reliability measure (BB). The number of iterations necessary to extract the values is another indication of whether or not the model capitalized on chance (Dijkstra, 1983). Each one of these measures gives slightly different information about the model as it reflects a "better" or ”worse" fit of the model with the data as described below. The R-square indicates how much of the variable is influenced by the information in the model. The chi-square should go down with each trimmed model indicating better fit between the data and model. The difference in the measure without and with the model should also be examined for a large decrease in in value when the model is specified (Bentler and Bonett, 1980). On the other hand, the BB reliability measure should go up with each new model if the model is showing a better fit with the data. Finally, the number of iterations is reported simply as a measure of the difficulty which the program had in extracting the values. So, I will examine the number of iterations across the models to see if they are consistent with each other, being aware that a large number of 184 iterations could indicate instability of the estimates. The information on each of these measures will be provided for each model's versions to assess the improved or decreased fit with the data. CHAPTER VIII FINDINGS I will begin by examining the results of the soft path preferences scale analysis. First the solar and nonsolar samples' scores will be compared to examine the first hypothesis of this project. Then the passive and active technology owners will be compared to determine if there are differences associated with the types of solar energy technologies. Finally, I will examine the results for subgroups of four different ownership lengths. In the second part of this chapter, the results of the modeling effort are presented and discussed. SOFT PATH PREFERENCES SCALE Soft path preferences scales were developed based on principal components analysis of the samples as presented in Chapter VI. SPP Scales: Solar - Nonsolar Homeowners: The first hypothesis explored with the SPP scale is H1:There is a greater amount of SPP among solar homeowners than among nonsolar homeowners. The SPP scale scores for the two groups are presented in Table 14 and Figure 5. 185 186 Table 14 SPP SCALE COMPARISON: SOLAR - NONSOLAR SoIar Nonsolar SPP Dimensions Mean s.d. Mean s.d. (n = 3809) (n = 2033) Self-reliance I. Energy 2.198 .706 2.964 .748 2. LocaI community 1.895 .379 1.512 .291 Natural resources conservation 3. Attitudes 1.904 .468 1.826 .356 4. Behaviors .990 .361 .542 .217 5. Energy conservation .492 .272 .745 .324 Personal gain 6. Noneconomic .623 .292 .729 .237 7. Economic 2.418 .539 2.936 .653 Total SPP Score 10.520 1.820 11.276 1.960 Average Solar [::] SPP score NonsoIar ngfl 3.0 _11 2.7 \ Q \ 2.4 ~1~L ‘- 2.1 \ c N \ o \ 1.2 \ 5‘ \ E \ “~..E rs. ‘TN o ‘\~ 0.9 \ EN 8 m g E u\ 0.5 \ 8\ g\ s Q 2 s\ SF\\. \\\ 34“~ w-‘\‘ “‘4 2“\. 0'3 $"\\.7;“\~ 13‘\~. : L OjN‘w c3‘\e C. U +4 .C 1.) C U 0.0 ul“\\ :3 <: ‘g 15 :2 LU‘\\ SéIf- Natural Persona] Reliance Resources Gain Conservation Figure 5: SPP SCALE COMPARISON: SOLAR - NONSOLAR 187 No test of significance of difference between means was conducted, because the solar sample was not randomly selected and because the sample sizes are so large. However the comparative differences in the means and standard deviations of the two samples will be discussed in an attempt to understand the soft path preferences of solar and conventional energy homeowners. An examinination of Table 14, indicates that in general the nonsolar homeowners have greater degree of soft path preferences than do the solar homeowners. The nonsolar homeowners have a mean score of 11.276, .756 points higher than the total score of 10.520 of the solar homeowners. This is contrary to hypothesis one that solar homeowners would have greater soft path preferences overall. The dimensions of the soft path preferences provide additional information about areas of agreement and disagreement between the two samples. The dimension of self-reliance indicates that the nonsolar homeowners show greater preferences for energy self reliance, but the solar homeowners have greater preferences for the local community's being the primary one responsible for decisions affecting its own development. The standard deviations Of the two samples are fairly similar, although solar homeowners show more variation regarding local community self reliance. 188 Larger differences are shown in the attitudes and behaviors about natural resources conservation. The solar homeowners show greater natural resources conservation attitudes (1.904 vs. 1.826) and behaviors (.990 vs. .542) while the nonsolar homeowners indicate having greater energy conservation behaviors (.745 vs. .492). There is a smaller difference in the expressed attitudes (.078) than in the natural resource conservation behaviors (.448 vs. .253). Issues of personal gain are clearly more important for the nonsolar homeowners than for the solar homeowners probably understandable, given the higher socioeconomic level of the solars. They placed more importance on the economic and on the status and comfort advantages than did the solar homeowners. The pattern which appears among the dimensions of soft path preferences suggests the nonsolar homeowners have greater preferences for energy self-reliance and economic personal gains. These are the dimensions showing the greatest differences between the samples. On the other hand, these are dimensions in which the solar homeowners experience less vulnerability since they have already implemented measures which make them more energy self-reliant. They also have greater income resources than do the nonsolar respondents. The next two areas of greatest differences between the two 189 samples indicate higher scores for the solar homeowners. They show greater preferences for local community self- reliance and broader natural resources conservation behaviors. These are perhaps linked to their greater involvement with community issues (Appendix C, Table 41). The natural resources conservation attitudes and preferences for noneconomic personal gains are practically identical for both groups: the energy conservation behaviors also show only small differences. SPP Scales: Active - Passive Technology Owners: As discussed previously, passive technologies have characteristics more consistent with the Lovins' soft energy technology notion than do the active solar systems. Therefore a second hypothesis is H3:Passive solar energy technology owners have more soft path preferences than active solar energy technology owners. To examine this hypothesis, the solar homeowners were divided into three types of solar technology owners: (1) owners of mixed active and passive systems (32%, n=1205), (2) active technology owners (60%, n-2278) and (3) passive system owners (9%, n-326). To keep the influence of the type of technology as clear as possible, the weights used in computing the SPP scales were 190 taken from principal components analysis of the separate groups rather than using the loadings developed for solar homeowners (Appendix C, Table 40). The results of the soft path preferences scale for active and passive solar technology owners are presented in Table 15 and Figure 6. Table 15 Figure 6 Consistent with the hypothesis, passive solar technology owners have a higher degree of overall soft path preferences (10.682) than do the active solar technology owners (10.109). The difference in the mean scores of .573 is in fact smaller than that found between the solar and nonsolar homeowners. Examining the separate dimensions of the scale more closely, there is a great difference (.762) in the active and passive preferences for local community self-reliance. The differences in preferences for energy self-reliance is smaller (.200) with the passive solar technology owners showing a slightly greater preference than the active technology owners. The natural resources conservation attitudes and behaviors are similar except for energy conservation behaviors. The active solar technology owners have higher scores on general 191 Table 15 SPP SCALE COMPARISON: ACTIVE AND PASSIVE OWNERS Active Passive SPP Dimensions Mean s.d. Mean s.d. (n = 2278) (n = 326) Self-reliance 1. Energy 2.122 .702 2.399 .656 2. Local community 1.774 .357 1.012 .173 Natural resources conservation 3. Attitudes 1.832 .451 1.616 .389 4. Behaviors .922 .333 1.163 .394 5. Energy conservation .467 .283 1.565 .512 Personal gain 6. Nonecomic .625 .314 .727 .257 7. Economic 2.367 .519 2.201 .494 Total SPP Score 10.109 1.748 10.682 1.649 Average SPP Score Active[::] Passive- 3.0 2.7 2.4 2.] s1\ 1.8 g \ 1.5 E:: 3E E:: >1 > !.2 P‘s :3 “‘4 3“‘\- “\ \ i‘\ \ N m\ \ 0.9 \.\I §L\\, -§ 'n~1‘ Egslfl .0 Juslp 0,6 “\ 8\ 3\ SN \ Ex E\ E%\‘ ,_N\~ I3‘\+ §;‘\~ E?‘\g u“\. g‘\\ 0.3 g\ 8\ +4\ 2\__m\ g\ 8\ 0.0 m\ 3\ <\ £\ LEI\ §\ m\ SeIf- ' Natura ersonal Reliance Resources Gain Conservation Figure 6: SPP SCALE COMPARISON: ACTIVE - PASSIVE 192 attitudes toward natural resources conservation, and fewer natural resources conservation behaviors. However, the passive solar technology owners report a greater number and variation in their energy conserving behaviors than do the active (difference - 1.098). Both the passive and active technology owners are similar in their personal gains dimensions. The economic advantages aspects of solar energy technologies are important issues to both solar samples, but less important for either of these groups than for the nonsolar homeowners. Solar Homeowners - Effects of Length of Ownership: Only a very small percentage of the American population owned solar energy technologies in 1980. In Chapter IV, I proposed that with increased length of ownership, solar homeowners will develop a network supportive of soft path preferences through association with others who own the technologies. Thus the hypothesis: H4:Length of ownership of solar technologies is positively associated with degree of soft path preferences. In order to examine this hypothesis, the solar homeowners were divided into four subgroups based on the number of years of technology ownership in 1980: (l) ten years or more (n8116); (2) 5-9.99 years (n=285), (3) 1-4.99 years (n=2728), and (4) less than one year (n=680). 193 Rogers (1983) claims that many new technologies fail within four years. It is clear that the majority of the solar homeowners (90%) were still very new adopters and hadn't passed this critical point. On the other hand, 116 claimed to have owned solar systems for a "long“ time period - e.g. before the "energy crisis" became very salient. About 8% (n-285) could be considered "old timers", having acquired the systems during the early part of the "energy crisis" decade of 1970's. Table 16 and Figure 7 present the soft path preference scores of the major subgroups. Table 16 Figure 7 The lowest soft path preferences score is 9.553 associated with those who have owned solar technologies for ten years or longer. The highest total score is associated with those who have owned the technologies for a year or less. There is more homogeneity in that subgroup while the subsample which has owned the technologies for the longest period has the greatest diversity of opinion as indicated by the standard deviations of 2.032 compared to 1.731. The relationship between SPP and ownership length is monotonic but in the opposite direction from the expectation of hypothesis four. Although the greatest difference in the SPP scale scores exists between those who have the technologies for less than a year and those who have owned them for more than 10 years 194 me.F mum.op NNm.F mm~.op mm~.~ mno.op mmo.m mmm.m mcoum mam Pouch woe. me¢.~ mew. mum.m one. moo.m mmq. Poe._ u_eo:oou .5 e—m. mac. mop. «em. «mm. was. 5mm. mms. u_eo:ooocoz .c :_mm Pecomcma Nam. mum. mum. mmo. mmm. mo“. Pme. MNo. .>cwm:oo zmcmcu .m moe. mmo.p ape. po~.P mac. mpm.~ Nae. mum._ Lo_>m;om .e mme. opo.~ 5mm. mom.p ~_e. cmo.. ume. omm._ mmcauwuu< .m :owum>gmmcoo mmuczommg Possum: mum. mom.p eom. «mm.~ mwm. __¢.F opm. mmm._ xuwczssoo Paco; .N one. mm~.N mmw. qpn.m mo“. ¢~N.N emu. mum.~ Augmcm .F mocmP—mguwpmm Acme u av Ammam - av imam n Ev Ae._ u as .e.m cam: .c.m coo: .e.m cam: .e.m cam: .ca p cog» mmmm .mg» eu_ .me» m-m .mtz + o_ amzmmmzzo do :Hozm; ”zomumazom :omue>eom:ou xusoem Low>mzom :omum>uom:ou mcwau ufieocoum oucawmom u~om ~aoo4 m:_mo umeocouocoz moczuwuu< cowua>somcou ou:a__o¢ ugom Amsocm w20_Hm mmca>mm xmeocm pouuoexm usoeezw aaumccoob mommo_o:;ooe mcwcc< cowum:_m>m “newczuoh eoumxm o>ummme mo comumpcou soumxm o>muo< mo :o«u_c=ou :owue03< nauseouom cowums_m>m ¢-:. oexb mo—ocsuo. meounohe mcmumsoeo >uo~o=guoh mo umou po__mum:_ o>mmmme use» co__mum:~ o>wau< use» E’s. cameo; magmuoczo ”a mczuqm mumou sauce; owmuo>< :m_mmbu Augean: 1o mounds. maxh mcd_~oza :m_m_LJ smoocmz mo chance 7. om< couc~_:u mo .oz cask v_o;mm:o: mpooz Ausocm moUL20mom Lesuxoucou l.,- .a goose: ezosu Lm_om “sodesm ~m:0mcom oeoo:_ camuausvm ascensm ~muouom usoeesm oumum uuoeesm ~muo4 mammogoczuoh «capawsm msogcosm oneoom :u_3 meadnoue moguMLoe Hmuoa an“: msoficosa e: mumou-mso~noce 2C”? mzc_b< :mqmmsu xmmocu: Co muuaae_ sex? u:_-oza :mmeuu xmeocm: mo muses; ou< cohsfimcu mo .02 make pmonomso: mpooz xuuocm moucosououe some uuom coflumsam>m mto_>mcom compa>somcou xmuocm msow>agom :omua>somcou mcmao amsocoum umou who: and on mmocucmgmmz mucm>wm Autocm couuoexm mo_mo_oc:uoh mcwpu< ouemmnom u~om neuoq comum3~m>m gmumcguoh mcweu omsocouosoz :omueop< ccossouoz mocsumuu< comum>somcou co_um:~m>m ~mpo=oo ouce_~o¢ «mom Ausocw /////IIIIIIII If!!! moussomom ~msuxoucou usoemsw ~acomuoe osoocm somumuscm ueoeesw queens; sconesm oueum 210 Table 19 THEORETICAL MODEL'S RELATIONSHIPS and GOODNESS OF FIT: SOLAR AND NONSOLAR HOMEOWNERS Theoretical Model Solar Baseline Model Latent Variables: Nonsolar Solar BETAS Needs - Resources .461 .415 Needs - Resources .386 Needs - Evaluation .010 .073 Needs - Tech Type -.067 Needs - Spp .160 .029 Needs - Spp .019 Resources - Eval .450 -.081 Resources - Tech -.391 Resources - Spp .065 -.153 Resources - Spp -.157 Tech - Evaluation -.122 Evaluation - Spp .335 .349 Evaluation - Spp .368 R-SQUARE Resources .213 .172 .149 Evaluation .207 .007 .015 Tech Type ---- ---- .177 Spp .195 .149 .161 No. parameters 31 31 45 No. iterations 9 18 86 Chi-square (no model) 5.141 3.628 6.725 (d.f) 300 300 703 Chi-square (with model) 2.309 2.013 4.909 (d.f.) 269 269 655 Bentler-Bonnett .551 .445 .265 211 also not related to the type of solar system (-.067) in the theoretical solar model. The nonsolar model indicates a small relationship between needs and SPP (.160). The contextual resources latent variable has a moderate (.450) relationship to the evaluation of the solar systems for the nonsolar sample, indicating that those with more resources tend to have a more positive assessment of solar systems. This relationship does not hold for the solar baseline model (-.O81). The solar theoretical model shows a moderate negative relationship between resources and technology type (-.391) suggesting that those who have more resources tend to have solar systems which have been more recently acquired, cost more, and have fewer operating problems. Both solar models indicate a weak but negative relationship between resources and soft path preferences suggesting that those with fewer contextual resources will have preferences more consistent with soft path changes. This finding is consistent with Lovins' interest in having the general populace rather than the elites change the energy base of society. The nonsolar homeowners' model shows no relationship (.065) between resources and soft path preferences. The evaluation of solar systems is moderate in its association with the development of soft path preferences in all of the models. 212 The model is not very good at capturing the R-square values for any of the variables. The variances of resources (.213: baseline - .172, theoretical - .142) and soft path preferences (.195; baseline - .149, theoretical - .161) are influenced to a low degree by the model. The evaluation latent variable is better explained for the nonsolar (.207) than the solar sample (baseline - .007, theoretical - .015). The R—square is highest for the technology type latent variable (.177) in the solar theoretical model. This and the negative and low (-.122) beta value between technology type and evaluation suggest that the use of solar energy technologies has a negative influence on the development of SPP. However, the evaluation latent variable is sufficiently strong to translate this negative input into a positive influence on SPP development. Therefore, SPP receives a stronger negative direct influence (-.157) from contextual resources than from technology type, contrary to expectations. But the technology influence is stronger than the negligible (.019) energy vulnerability needs influence as expected. Evaluation of the solar technologies, however, does have a moderate (.335: baseline - .349, theoretical - .368) influence on the development of the soft path preferences. This is further explored through trimming the original theoretical models. 213 The models were trimmed by dropping one of the latent variable relationships at a time based on the theoretical model's beta values until a linear relationship was portrayed. First the relationship between needs --> evaluation or needs --> technology type was dropped for all three models: then the relationship between needs and soft path preferences was constrained to 0: finally, the relationship between resources and SPP was constrained leaving a straight linear relationship from needs --> resources --> technology type or evaluation (with technology type --> evaluation included only in the solar theoretical model) --> SPP. With each additional constrained latent variable relationship, there was a slight improvement in the fit of the models until the best fit was shown with the linear form of all models (Appendix C, Table 45). The strongest fit appeared with the nonsolar model going from BB-.570 (dropping needs --> evaluation) to BB-.582 (straight linear model). The solar theoretical model goes from BB-.258 (dropping needs --> tech type) to BB - .270. Both show only the most trivial minimal improvements in BB score of .018 (solar) and .012 (nonsolar). The nonsolar model extracted its values in only 8 or 9 iterations, while the solar theoretical model required 64 to 93 iterations. The other criteria for indicating better fit of the model with the data did not improve to any important degree. 214 Theoretical Model: Active and Passive Technology Owners: Given the better fit which is shown by the nonsolar than the solar model, the next attempt was to model the relationships among the passive and active solar technology owners in order to see if the relationships among the latent variables were different for those subgroups. The manifest variables relationships to the latent variables are indicated in Table 20. Table 20 Although the relationships between the manifest and latent variables for the passive and active technology owners were very similar to those of the solar sample, there are some differences. Age continues to have a moderate influence on perceptions of energy vulnerability and needs for both the passive (.426) and active (.575) solar system owners. Number of children has a low negative influence on the passive owners (-.203) and no influence on the pure active owners. On the other hand, the type of household has a moderate influence (.506) on the active solar system owners' perceptions of energy vulnerability needs while it has no (.060) influence on the passive homeowners. The physical characteristics of the house (.267 vs. .097) and 215 Table 20 THEORETICAL MODELS' MANIFEST VARIABLES: ACTIVE AND PASSIVE TECHNOLOGY OWNERS Variables: Active Passive (Weights) ENERGY VULNERABILITY NEEDS Household Type .506 .060 Number of Children -.024 -.203 Age .575 .426 Dwelling Type .097 .267 Energy Costs .312 .784 Impacts of "Energy Crisis” .035 -.217 Future of ”Energy Crisis" .088 .066 CONTEXTUAL RESOURCES Income .638 .932 Education .005 -.073 Federal Support .569 -.052 State Support .146 -.035 Local Support -.037 -.101 Solar Group Member -.027 -.310 Problems - Costs Up -.026 .112 Problems With Local Policies -.018 .106 Problems With People .109 -.087 Problems Building Technologies -.178 -.026 TECHNOLOGY TYPE Ownership Length .874 .307 Year Active Installed .767 ---- Year Passive Installed ---- .119 Cost of Technologies .388 .680 Active Building Method .418 ---- Passive Building Method ---- .848 EVALUATION General Evaluation .742 .726 Technical Evaluation -.041 .124 Expected Energy Savings -.451 -.070 Adding Technologies .685 .598 Recommend Adoption .723 .752 Condition of Active System .404 ---- Condition of Passive System ---- .320 Technical Support .357 .180 SOFT PATH PREFERENCES Energy Self Reliance .856 .829 Local Self Reliance .614 .648 Conservation Attitudes .636 .576 Conservation Behavior .338 .455 Energy Conservation Behaviors -.l95 -.325 Economic Gains .586 .561 Noneconomic Gains .576 .506 216 the cost of energy seem to be stronger influences on passive technology owners' perceptions of energy vulnerability needs. The average cost of heating the water and heating and cooling the home has about twice as strong an influence on the passive owners' needs perceptions (.784) as compared to the active owners (.312). The future of the energy situation in the nation has no impact on either type of system owner, while the perception of impacts on household's life style is negatively related to the passive owners' needs perception (-.217). As before, income is the manifest variable which dominates (active - .638, passive - .932) the resources latent variable. Again, education (active - .005, passive - -.073) does not influence the perception of contextual resources. Another variable which is a moderately strong (.569) resource for active solar homeowners is federal subsidies. State support also provides some resources to the active solar homeowners (.146). Local support was noninfluential (-.037) with active owners and had a small negative influence (-.101) on the passive homeowners' resources. The second largest influence on the passive homeowners' contextual resources is participation in groups which support solar energy, although the influence is negative and moderate (-.310). 217 The various problems in acquiring solar systems play a negligible influence on the perception of contextual resources. Problems experienced with people's reactions to the technologies and with acquiring the solar systems were stronger influences on the active latent variable while cost factors and problems with local building codes and access rights were greater issues for passive solar owners. The length of ownership is more influential on defining technology type for the active system owners than for the passive system owners. Both manifest variables which measure this have very strong loadings on the technology type latent variable (.876, .767). On the other hand, the cost of the technology (.680) and the method used to build it (.848) were stronger influences on the technology type of the passive owners. As seen previously, the general evaluation variable (active a .742, passive - .726), the plan to add solar systems (active - .685, passive i .598) and the recommendation to others for adoption (active - .723, passive - .752) have the strongest influences on the evaluation latent variable. The evaluation of the specific technical capabilities of the solar systems has practically no influence on the active solar homeowners, while it does have a small (.124) effect on the passive solar systems' evaluation. 218 There is a great difference in the influence of the expectations of savings on the solar active and passive system owners. These expectations have no influence on the passive solar technology owners' evaluation while they have a moderately strong (-.451) negative influence on the active technology owners. Perhaps the passive owners did not have expectations for great savings, or they were more accurate in their estimation of what they would save than the active solar homeowners. The technology support structures are more influential on the active system owners (.357) than the passive (.180). The assessment of the condition of their technologies has approximately the same effect on the evaluation latent variable for both active (.404) and passive (.320) system owners . Finally, the subsamples' soft path preferences are very similar to each other and to the overall solar sample. Again, the diversity which does occur is in relation to the attitudes and behaviors about natural resources conservation. The active solar homeowners record a stronger influence of natural resources conservation attitudes (active - .636, passive - .576) while the passive solar system owners show greater general conservation behavior (active - .455, passive - .338). The number of energy conserving behaviors which they report has a low moderate 219 negative (-.325) influence on the soft path preferences of passive solar system owners and it has a small, but still negative (-.195) influence on the soft path preferences of the active solar system owners. This is consistent with the solar sample. The relationships among the latent variables of the model are also important to examine for both the active and the passive solar homeowners. Table 21 In this model, as in the solar model, the highest beta weights are between needs and resources (passive - .430, active - .435). The relationship between evaluation and soft path preferences is similar, moderately strong. The other sizeable relationship between the latent variables is between resources and technology type showing a moderate negative relationship (active--.366, passive--.432) as was found in the total solar sample. The other relationships are negligible for the active system owners. However the passive owners' model has a positive and small relationship between technology type and evaluation (.214), and between energy needs and technology type (.139) The models' R-squares indicate that the model influences the 220 Table 21 THEORETICAL MODELS' RELATIONSHIPS AND GOODNESS OF FIT: ACTIVE AND PASSIVE TECHNOLOGY OWNERS Latent Variables: Active Passive BETAS Needs - Resources .435 .430 Needs - Tech Type -.088 .139 Needs - Spp .066 -.129 Resources - Tech -.366 -.432 Resources - Spp -.056 -.155 Tech - Evaluation .040 .214 Evaluation - Spp .319 .330 R-SQUARE Resources .189 .185 Tech Type .170 .154 Evaluation .002 .042 Spp .107 .195 No. parameters 42 42 No. iterations 12 13 Chi-square (no model) 6.142 7.198 (d.f.) 595 595 Chi-square (with model) 3.525 6.057 (d.f.) 550 550 Benter-Bonnett .426 .159 221 greatest percentage of the variance of the soft path preferences for the passive system owners (.195) while it captures less of the actives' SPP (.107). The strongest latent variables are resources (passive-.185, active-.189) and technology type (passive - .154:active - .170). The models capture very little of the variance of evaluation for either the passive (.046) or the active (.002) system owners 0 Comparing all three models - solar, active and passive system owners - evaluation is the latent variable least affected by the model. The passive solar technology owners experiences with their technologies positively affect (.214) their evaluation of the solar technologies. The technologies also show a postive relationship with the development (.330) of their soft path preferences. This is not true for the active solar energy technology owners. Therefore hypothesis two, that technology will have a greater effect on the development of soft path preferences than contextual resources and energy vulnerability needs is supported for the passive solar technology owners but not for the active owners. The theoretical models were further trimmed to see if a better fit could be obtained between the models and the data. In order to do this, the relationship between needs --> technology type was constrained to zero. Next, 222 resources --> SPP was dropped, then needs --> SPP was constrained making the models completely linear (Appendix C, Table 45). The linear relationship provided the best improvement in goodness of fit for the passive model (BB-.162 to BB=.171) while the active model remained very similar (BB-.422 to BB=.423) with the non-trimmed model capturing slightly more of the active (.426) data patterns. The number of iterations were very similar for both models. Trimming strengthened all the latent variable relationships' in the passive model, while the active model remained basically the same (Appendix C, Table 46). Modified Model: Since the technology problems variables contributed so little to the contextual resources latent variable I decided to create a latent variable composed only of five technology problems. I then placed it between technology type and evaluation in the model therefore modifying the original theoretical relationships. This is shown in Figure 10. Figure 10 As indicated in Table 22, the modified SPP development models are composed of 33 manifest variables for the solar sample and for the passive subsample: the active subsample 223) som>mzom eomue>somcou xueocm Lee>asom somuu>uomcou mcueo uMsocoum oucmm_o¢ u~om ~auoa mcmmo oesocouocoz mocsumuu< :owum>som:ou oucmm—oa «.om Amhocm moucosououe somums~m>m msodnoee sued uuom mo_mo_ocsuoh mcwpu< :omues~m>m ~e0ucguoh macumxm o>mmmme Locomumpcou msoumxm o>muu< mo comuwpcou :omueop< pcossouom :omua:~a>m gahosoo mmcm>nm xmsocm pauuoexm ascensm n¢Umczuoh ,;:, mzc_H< :m_mmsu xmsocm: mo mouse:— oexh mcmumoza :mmmmuu smsocm: mo spouse om< cospfimzu mo .02 oexh p—ozomsoz \. mpooz meoflnosm «embasoco xmsocm modmo_o::uob mcmpnmsm mso_noue 0.900; sum: men—nose mo_u_~om ~auom sums msounose a: mumou-msoflaome a some >uo~0=zuoh eo—umumc_ o>~mmae umo> voHHaumc_ o>muu< sao> sauce; emgmuoczo xmo—ocsuoh mo «moo vogue: meme—mam o>euo< / e93»: 5:315 2:33 / MQUHSOmflm ~0=HXOH=OU .I/‘III O I'll-‘IIII‘I‘I. .1 '1... till Loose: macho LaLom usoeesm ~acomeoe meow:— usoeesm ~muopou sconesm venom 224 uses 30 manifest variables in its modified model. The six latent variables follow the same modes as specified previously in Table 13. Table 22 The manifest variables which were omitted from the complete solar sample and the active samples due to their having low values were: number of children, future of the "energy crisis", impacts of the "energy crisis", local support, personal support, solar group member, and education. The passive sample model ommitted the following: household type, local support, state support, and education. As seen previously, the active technology owners and solar samples are very similar. The values of the manifest variables are also similar to those of the theoretical models. The primary difference is in the problems latent variable. Its manifest variables have a positive and much stronger relationship with the latent variable indicating that they are better represented as a separate variable. The influence which the manifest variables have on the problems latent variable is similar for all three samples. The strongest influence is created by the problems which the solar system owner has in operating the solar system (solar-.920, active-.915, and passive=.513). The second major effect is created by problems encountered while trying 225 Table 22 MODIFIED MODELS' MANIFEST VARIABLES: SOLAR, ACTIVE AND PASSIVE HOMEOWNERS Variables: Solar Active Passive (weights) ENERGY VULNERABILITY NEEDS Household Type .508 .494 ---- No. of Children ---- ---- -.143 Age .513 .595 .469 Dwelling Type .190 .102 .285 Energy Costs .399 .312 .761 Impacts of "Energy Crisis" ---- ---- -.l98 Future ”Energy Crisis” ---- ---- .109 CONTEXTUAL RESOURCES Income .706 .666 .903 Education ---- ---- ---- Federal Support .543 .578 .018 State Support .093 .138 ---- Local Support ---- ---- ---- Solar Group Member ---- ---- -.313 TECHNOLOGY TYPE Ownership Length .698 .881 .220 Year Active Installed .661 .776 ---- Year Passive Installed .267 ---- .083 Cost of Technologies .308 .380 .478 Active Building Method .449 .397 ---- Passive Building Method -.114 ---- .701 PROBLEMS Problems - Costs Up .332 .385 -.075 Problems With Local Policies .259 .334 .133 Problems With People .332 .387 .285 Problems Building Technologies .685 .632 .498 Operating Problems .920 .915 .513 EVALUATION General Evaluation .832 .834 .430 Recommend Adoption .758 .766 .382 Condition of Active System .620 .714 ---- Condition of Passive System .337 ---- .321 Technical Evaluation .268 .232 .089 Adding Technologies .355 .286 .185 Technical Support .378 .382 .162 Expected Energy Savings -.507 -.554 -.131 SOFT PATH PREFERENCES Energy Self Reliance .850 .860 .358 Conservation Attitudes .582 .590 .154 Noneconomic Gains .596 .524 .244 Local Self Reliance .679 .599 .335 Economic Gains .536 .712 .251 Conservation Behavior .382 .179 .218 Energy Conservation Behavior -.231 -.l42 -.104 226 to build the system (solar=.685, active-.632, passive=.498). The other problems have a low moderate influence on the solar and active owners. Problems with increased costs have no influence on the passive owners' latent variable (-.075), problems with local codes and solar access were also a minimal issue (.133) for the passive solar owners. The six latent variables model shows an improvement in fit for the subsamples but not for the solar sample as seen in Table 23. Table 23 The full solar sample goes from BB-.215 with the full modi- fied model, to .214 with the linear trimmed six latent varia- ble model (Appendix C, Tables 47 and 48). This is a worse fit than the 88-.270 of the theoretical model for the solar sample. The active solar technology owners, however, show an improvement in fit with 88-.469 with the full modified model to .480 with the linear trimmed one. This is better than the .426 value of the full theoretical model. Again the passive owners' subsample shows a different pattern than the active and solar samples. Its best fit is with the linear modified model showing BB-.293 compared to .240 with the full modified model. This shows a much better fit than the BB-.159 of the full theoretical model. 227 Table 23 MODIFIED MODELS' RELATIONSHIPS AND GOODNESS OF FIT: SOLAR, ACTIVE AND PASSIVE HOMEOWNERS Variables: Solar Active Passive BETA's Needs - Resources .434 .427 .434 Needs - Tech Type -.l78 -.102 .117 Needs - Spp -.099 .117 -.099 Resources - Tech -.472 -.347 -.474 Resources - Spp -.l96 -.079 -.190 Tech Type - Problems .002 .043 -.002 Problems - Evaluation .259 .373 .259 Evaluation - Spp .203 .187 .206 R-SQUARE Resources .188 .182 .188 Tech Type .182 .161 .183 Problems .000 .002 .000 Evaluation .067 .139 .067 Spp .129 .045 .126 No. parameters 45 38 41 No. iterations 15 ll 16 Chi-square (no model) 7.737 5.096 6.494 (d.f) 666 435 528 Chi-square (with model) 6.078 2.705 4.937 (d.f.) 614 390 480 Bentler-Bonnett .215 .469 .240 228 In every sample, the R-square is nonexistent for the problems and very low for evaluation, while the best values for the latent variable soft path preferences is with the passive solar system owners modified model. This reflects the fact that few of the solar technology owners had problems with their technologies (see Table 42). The beta values have a familiar pattern. The strongest relationship continues to be between the needs and resources latent variables. The relationship between resources and technology type are also moderate for every sample (solarsa .472, active--.347) and for the passive owners, -.474. The relationship between technology type and problems are non-existent, yet the relationship between problems experienced and evaluation of the solar system is similarly low for all the samples, the solar and passive samples - .259 and active samples - .373. While the relationships between evaluation and soft path preferences is low for the active solar (.187), it is again more similar for the solar sample (.203) and for the passive owners (.206). Even with a better fit between the model and the data for the modified model, the patterns of relationships continue to be stable. The latent indicator of contextual resources 229 has a weak relationship to the development of soft path preferences for the passive sample and has no effect for the active sample. Energy vulnerability needs also does not have much of an effect at this historical period. Evaluation of the technology has a moderate influence on SPP development for all three samples although the type of technology has a low influence on evaluation. CHAPTER IX DISCUSSION AND CONCLUSIONS Since 1980 when the data were gathered for this project, there have been many changes in the national policy support for the development of renewable energy technologies. Concerns about economic development and growth have become more salient and concerns with energy scarcity have receded. Although the public does not now have the high awareness of energy resources as was present in 1980, there are other aspects of the environment which continue to capture concern about natural resources and their developments (Morrison, 1986). There are also those who warn that natural resources scarcity is still a relevant issue. Therefore lessons which we can learn from the "energy crisis" are still relevant to policymakers, planners, and activists of the future. Are there indications that there has been a change in lifestyle preferences which are more compatible with a renewable resources energy base? Many researchers have suggested that there has been a qualitative change in the core values of the American society during the 1970's (Mitchell, 1984; Yankelovich and Lefkowitz, 1980: Buss and Craik, 1983; Milbrath, 1984). 230 231 Are the worldviews (Buss and Craik, 1983) supportive of soft path structural changes associated with changes to a renewable energy base for society? More specifically, is the use of solar energy technologies associated with soft path preferences? Is the general population of society supporting this change, or is such a change dependent on political and economic elites? For theorists interested in understanding the interaction of energy, technology and society, these questions are also important. Although Amory Lovins is primarily an energy activist, his thinking has provided one of the clearer, though very limited, recent theories of social change created through the societal energy base. This has been the foundation for my exploration of the interaction of solar energy technologies with soft path compatible attitudes and behaviors of the American population in 1980. THEORETICAL IMPLICATIONS Theoretical Implications of the SPP Scale: Lovins has suggested that currently the societal values in the United States and other European countries are supportive of the development of the soft energy path. He suggests that the population is interested in: (1) self- reliance, to gain independence from energy utility companies and from other forms of centralized energy control in order to take control over the energy aspects of their lives. 232 (2) energy conservation, so that there will be more available energy resources for future generations and for other nations; (3) personal gains of more community cohesion, self-esteem (e.g. lack of alienation), and satisfaction with life. He suggests that these preferences can be accomplished through maximizing economic and noneconomic benefits for individuals and communitites. As households, individuals and communities accomplish these goals, they will be restructuring society into a more benign system with a more resilient, less catastrophic, and equitable form. Because he is concerned that this be accomplished through a changed energy base, his primary focus has been on encouraging people who use traditional energy forms to switch to renewable energies and to conserve energy resources . Although Lovins is clear that he expects the values which are supportive of the soft energy path to already be in place, he is not clear about what the specific impacts of the implementation of those values through renewable energy technologies and conservation will be, except through a restructured social system. He has argued that the criteria used in choosing an energy technology should be its end use. Therefore he suggests that it is appropriate to use ”soft" as well as "hard" energy technologies during the transition period before the social system is restructured 233 to be more compatible with the "soft" energy technologies. I have suggested that the development of the soft energy path social changes will not occurr simply through the use of renewable energy technologies and energy conservation. These must reinforce and be guided by values, attitudes and behaviors which are compatible with a "soft" restructuring of society. Otherwise there will be a continuation of the current hard energy path with those in control using the surplus natural resources created through conservation and a change in energy base for their own purposes (e.g. centralized political control and economic production). Therefore, I have proposed that the change in energy base must be associated with compatible attitudes and behaviors. The Lovinses have increasingly stressed the economic advantages to individuals and communities of changing their energy base. I suggest that economic gains are not a sufficient reason for using soft technologies. A stress on economic gains will further reinforce the hard energy path. Noneconomic concerns for self-reliance and more equitable distribution and control of natural and socioeconomic resources must occur. Furthermore, I have argued that it is necessary for compatible attitudes and behaviors to persist over time 234 among the renewable energy technology owners. Otherwise the soft path will not overcome the resistance of the hard energy path. Those who change their solar energy base will lead the change to the soft path if the technologies reinforce and increase the values, attitudes and behaviors which are compatible with such social change. This should be true if renewable energy technologies serve as mode of expressing those values. The soft path preferences scale was developed to measure the attitudes and behaviors of those who changed their energy base compared to those who had not done so. The soft path preferences scale is composed of dimensions which are compatible with a soft path restructuring of society. These dimensions include: (1) self-reliance, (2) natural resources conservation, and (3) personal gains. The scale has ommitted the important dimensions of (4) equity and (5) social diversity due to limitations of the data available for the present study. The scale was then used to examine the relationship between the ownership of solar energy technologies and attitudes and behaviors compatible with soft path social changes. The results of this examination are the following: (1) Nonsolar homeowners have more attitudes and behaviors which 235 are compatible with the development of the soft path social changes than solar homeowners. (2) Solar homeowners who have owned the technologies for a shorter period of time express more attitudes and behaviors which are compatible with the development of soft path social change than do those who have owned the technologies for a longer period of time. (3) The passive solar technologies owners have more attitudes and behaviors which are compatible with the development of the soft path social change than do the active solar technologies owners. Only the last of these findings supports the Lovins' notion as developed in my hypotheses. The fact that the soft path preferences scores are higher for the nonsolar homeowners than for the solar homeowners does, however, indicate that Lovins is correct in proposing that the attitudes and behaviors held by Americans are supportive of aspects of the development of the soft energy path social change. There is broad based support for the decentralized, self-reliant, energy conserving behaviors which he has proposed. It is not clear that such support extends to using solar energy technologies thus changing the energy base. The lower SPP scale scores of the solar technology owners suggest that those who have the greatest attitudinal and behavioral support for the soft energy path (the nonsolars) 236 are not changing the energy base of their household. This finding was explored through a comparison of active and passive solar technology owners. In this comparison, the passive solar technology owners have a higher soft path preference score than the active owners. Since the solar sample is dominated by the active solar technology owners (60%), it is clear that their preferences overshadow the attitudes and behaviors expressed by the passive solar technology owners (9% of the solar sample). The passive solar energy technology owners' attitudes and behaviors are more compatible with the development of the soft energy path social change as seen in examining the dimensions of SPP. The passive solar technology owners were the main implementers of soft path social change in 1980. They perform far greater energy conserving and other natural resources conserving behaviors than do either the active solar technology owners or the nonsolar homeowners. As argued previously, their solar energy technologies are more ”soft" than the active technologies. They do express, however, the lowest attitudinal support for natural resources conservation of any of the samples. Their preferences are higher for noneconomic personal gains 237 but lower for economic gains than those expressed by the active technology owners. These SPP scores plus their strong concern about energy self-reliance suggest that the passive solar technology owners view a change in energy base as desirable based on concerns for resource conservation (Stern and Aronson, 1984) and self-reliance. This leads to the speculation that these technologies owners may have stronger concerns for greater equity across generations. They, however, do not show as many preferences for community self-reliance as the other samples do. Therefore this may indicate a preference for an individual approach to resource conservation rather than the cohesive community approach advocated by the Lovinses. This issue, of course, can only be examined with additional data. Another pattern is reflected in the SPP scale scores of the active technology owners and the nonsolar sample. The active owners' expressed natural resources conservation attitudes are similar to those of the nonsolar homeowners (1.832:l.826), while showing higher natural resources conservation behaviors (.922:.542) and lower energy conservation behaviors (.467:.745) than the nonsolar homeowners. They also showed a stronger preference for economic gains than the passive sample (2.367:2.201), yet lower scores than the nonsolar sample did (2.367:2.936). Their noneconomic personal gains preferences were also the lowest of the three samples, as was their concern with 238 energy self-reliance. These dimension scores suggest that energy is a commodity for the active solar technology owners rather than a tool congruent with soft path social change (also see Stern and Aronson, 1984). This orientation and the dominance of active solar technology owners may be the reason for the pattern of SPP dimensions which was observed over time among the solar owners . There is a negative relationship between time and the SPP scale scores among solar technology owners. As seen previously, those who have owned the solar energy technologies for five years or less seem to meet a watershed period close to the fifth year. At that time both the economic gains and the energy self-reliance concerns of the householders become stronger. Five years may be the time when the idea that the solar energy technologies are a commodity generating economic gains is challenged. This challenge is in conflict with the strong concern for energy self-reliance which appeared at this time in the sample. Those who have owned the solar technologies for five years or more begin to express preferences for more noneconomic gains from the technologies. 239 The nonsolar homeowners' SPP scale is dominated by concerns for energy self-reliance and economic personal gains. They show a preference for community self-reliance which is closer (1.512) to the stronger concern of the active solar onwers (1.774) than the weaker preferences of the passive solar technology owners (1.012). These findings suggest that the Lovinses' emphasis on economic personal and community gains of changing the energy base of society may be sufficient for nonsolar homeowners to change their energy base. However, if this course is followed, energy will be perceived primarily as a commodity to be incorporated by the current social structures. There will be no change to the soft energy path because the elites of the hard path will simply control this resource as they have other resources (Reece, 1979: Purdy, 1985: Schnaiberg, 1980). Furthermore, those who perceive energy as a commodity are also not strong energy conservers. These findings suggest that the Lovinses' current emphasis on energy as an economic gain and the emphasis on using both hard and soft technologies during the transition will not support a change to the soft energy path. These findings indicate that there is a continuation of the two technologies pattern first identified by Mumford (1967) 240 within solar energy technologies. One is a "democratic technology" (e.g. passive solar energy technologies) and one a "totalitarian technology" (e.g. active solar energy technologies) more dependent on elite dominated systems. Therefore, even with a society based on a renewable energy base, the separation of the technocratic elites and the populace can continue though perhaps not quite as strongly as the separation which exists within the current hard path. Once these differences between the nonsolar and solar samples, active and passive samples were discovered, then I further investigated the processes which support the development of the soft path preferences through the use of the SPP development model. Theoretical Implications of the Soft Path Preferences Model: Lovins has proposed that the implementation of the solar energy technologies will create more satisfaction, diversity, equity, and soft path structural changes. I have suggested that in order for these changes to occur, the solar energy technologies must reinforce attitudes and values which are compatible with the soft path. Lovins also suggests that the only reason that more people are currently not using solar energy technologies is because there are so many institutional barriers preventing the renewable energy base technologies from being economically 241 competitive. He suggests that the primary task of energy activists is to reduce those barriers. This issue is explored through the modeling process. The main finding is that although perceived energy vulnerability needs and contextual resources were able to moderately influence the solar energy technologies' characteristics, the technologies had little association with the problems which the solar technology owners encountered or with their evaluation of the technologies. Therefore the technologies had little influence on the development of the SPP. This was true for every sample examined except for the passive sample. The type of technology which is used by the passive solar technology owners has a stronger and positive influence on evaluation of the technology and on the development of the SPP. Since there was no relationship between technology type and problems and since the model did not explain any of the variance in the problems encountered by the solar energy technology owners, this finding casts doubt on the need to reduce barriers for early adopters (e.g. less than five years ownership length). The early adopters reported few barriers. Perhaps barriers will be encountered by later adopters, but those were not the concerns of these affluent early adopters. 242 At this early stage in the adoption process, the primary influence on the development of preferences compatible with soft path social changes was the evaluation of the solar energy technologies. Therefore, the symbolic value of the solar energy technologies dominated. The evaluations were generally positive irrespective of experiences with the technologies. This is seen in the nonsolar as well as in the solar samples. This, however, is different for the passive solar technology owners. The experiences which this sample had with the solar energy technologies did serve to more strongly reinforce the positive evaluation of the solar energy technologies and strengthened its effect on the development of the SPP. Only the nonsolar sample's model captured some of the variance (.221) of the evaluation variable suggesting that contextual resources do influence positive evaluation of solar energy technologies for nonsolar homeowners. The model does not capture any of the evaluation variable's variance in the solar samples. Therefore this needs further exploration. The symbolic nature of the solar energy technologies' influence in the development of SPP suggests why the solar sample's SPP scores are reduced over time. During five years of working with the solar energy technologies the owners are able to move past their symbolic nature and experience their 243 realities as well. Therefore, the technologies are not reinforcing the development of SPP but reducing it. Broader natural resources conservation behaviors slowly increase over time, however. Unfortunately, the historical natural resources conservation behaviors of the nonsolar sample is not available in these data to form a comparison group. Therefore these behaviors may simply reflect the preferences of a different historical cohort. Besides indicating the lack of influence of the solar technologies on the development of SPP, another result of the modeling indicates that income is the dominant resource for all the samples. Federal subsidies are a second resource for the active solar technology owners. Belonging to a solar profession or voluntary solar energy group has a negative influence (-.310) on the passive technologies owners resources. The resources are negatively related to soft energy technologies for all three samples. This suggests that those with many resources will select the "harder” solar energy technologies (e.g. active). Resources are positively related to evaluation of the solar energy technologies for the nonsolar homeowners. The evaluation variable of the nonsolar homeowners captures the symbolic nature of solar energy technologies since those homeowners report their perceptions without owning the 244 technologies. The dominant manifest variable is general favorability to solar energy technologies, although technical characteristics are also important. The second strongest manifest variables are the economic ones, as expected. Also, the personal networks have a greater influence on the nonsolar homeowners suggesting that this is a potentially important resource for the introduction of solar energy technologies to nonadopters. These findings support the possibility of there being a bifurcation in the renewable energy base of society. Those with more resources prefer the harder solar energy technologies while those households with fewer resources prefer softer renewable technologies. This relationship holds true across all solar samples. The solar samples are primarily composed of socioeconomic elites, and although they have adopted the solar energy technologies thus changing the energy base of their households, their greater resources are not supportive of soft solar energy technologies. On the other hand, the nonsolar sample which represents a cross-section of the population, does show a positive association between access to resources and greater SPP development. The models for the nonsolar homeowners and the passive technology owners capture about 15% of the SPP variable's 245 variance while the active sample's model contributes only about 10%. This leaves room for considerable theoretical exploration of other models of SPP development. The models do, however, gives some indication that the change in energy base proposed by Lovins through the use of solar energy technologies has little direct influence on the development of attitudes and behaviors compatible with soft energy path social changes in the short run (e.g. 1-5 years). Because of the early adopter nature of the solar samples, the long term direct effects cannot be measured with these data. Due to the household nature of the data, the indirect effects through social structural changes are also not measured. Therefore Lovins' claim that the changes to an alternative energy base will create social structural changes to the soft energy path is not supported here, but is not definitively tested. These findings further indicate that Lovins' theory of social change must be refined to include diversity in the impacts of the change in energy base of society depending on whether the change is based on hard or soft renewable energy technologies. The active solar energy technologies change in energy base will be incorporated into the hard energy path providing no challenge to the current social system. A 246 change in energy base which is produced through the use of soft renewable energy technologies does have an influence on the development of attitudes and behaviors which are more compatible with the development of social structural transformation to the soft energy path. Although Lovins has allowed for the use of many different types of renewable energy technologies, he prefers passive energy technologies for households. This is also the type which the Lovinses use in their home. POLICY AND PROGRAMMATIC IMPLICATIONS The SPP development model indicates some leverage points for those seeking to develop policies and programs supportive of a change in the energy base of society and for those seeking soft energy path social changes. One of the first decisions the policy-makers or energy activist must make is whether their focus will be simply on a change in the energy base or a change which is more compatible with attitudes and behaviors consistent with soft energy path changes. I will assume that the policy-maker is more interested in changing the energy base of society, while the energy activist is interest in restructuring society along the soft energy path. PolicyImplications: Since the primary purpose of the policy-maker is to change the energy base of society, the primary focus will be on how 247 to encourage nonsolar homeowners to conserve energy and to use active solar energy technologies. As indicated in the energy vulnerability needs relationships of the nonsolar homeonwers, the policy-makers should focus primarily on single family homeowners composed of middle aged (35-45 years old), two parent or multigenerational families with children, who earn high incomes. The types of nonsolar people who would be most amenable to solar energy technology adoption have already been thoroughly identified by Farhar-Pilgrim and Unseld (1982). Past policies supportive of solar adoptiont at the federal and state levels have little impact on the nonsolar homeowners. But the nonsolar respondents are concerned about the cost of the solar technologies as indicated by the .641 weight of their evaluation latent variable. Business incentives to solar energy technology companies for dependable and knowledgeable support for acquiring and operating the solar energy technologies would respond to another concern about use of solar technologies. Specific technical support such as warranty and insurance coverage which will reduce the risks of adopting the solar energy technologies are also important to the nonsolar homeowners (.364 on the evaluation variable: also see Table 27). The approach used in encouraging the use of solar energy 248 technologies should stress that the technologies will increase energy supply, providing greater independence from a need for big power plants and utilities as well as independence from national policies. Furthermore, the image should be one of self-reliant personal capabilities (see Table 34). Little mention should be made of the "energy crisis" and the need to decrease personal use of energy to help resolve the energy crisis. As seen in the energy vulnerability variables, this had little influence on household's energy vulnerability needs perceptions. The personal economic gains of solar energy systems should also be stressed. The suggestion that solar energy technologies will decrease utility bills and increase the home resale value are important to the nonsolar sample. However, it is critical not to stress great savings compared to current utility bills. As seen in the evaluation variables, when expectations of savings are too high, the experience of not meeting those expectations reduces the positive assessment of the technologies (-.451) for active solar owners. Finally, the SPP model provides some suggestions about the process to be followed in diffusing solar energy technologies. (This has been closely examined in Farhar- Pilgrim and Unseld, 1982). The emphasis should be placed on 249 reaching nonsolar homeonwers through their personal networks, for this is the manifest variables that showed the second largest weight (.386) on the contextual resources latent variable. These approaches will encourage a change in the energy base of society to hard renewable energy technologies, the active solar technologies. There will be no social structural transformation, but the social system will continue on its hard energy path. These approaches treat energy as a commodity. Implications for Energy Programs: Energy activists seeking to transform the social structure to the soft energy path will need to emphasize the soft solar energy technologies (e.g. the passive technologies) and energy conservation. The people approached should be similar to those of interest to the policy-makers. However, the energy activist should primarily emphasize an approach to those owning single family dwellings who have larger families, rather than emphasizing the two parent families. Time spent emphasizing policies supportive of solar energy technologies, at the national or state levels, will have little effect on the households interested in passive solar 250 energy technologies. However, this was true in 1980 when the policy incentives generally were not extended to passive technologieslo. Perhaps the influence is greater since the policies have become more favorable. This question needs further research. Greater emphasis should be placed on changing local building codes and in protecting solar access rights (.106 weight on the contextual resources) as well as on reducing the cost of using passive technologies due to increased difficulty in getting financing (9%, n-29 had this problem) or increased property taxes (8%, n-25). The potential passive solar energy adopters will be interested in having adequate technical support (.180 on the evaluation variable) and will want to know about the technical capabilities of the passive solar energy technologies (.124 on evaluation). They will also be interested in monitoring the performance of their solar energy technologies (.320). The process used to encourage passive solar technology use should emphasize the importance of the technical capacities of the passive solar energy systems to help the household become more energy self-reliant. It should also emphasize the positive aspects of being involved with the construction of the technology (.848 on technology latent variable) and 251 on the technologies' capacities to further increase natural resources conservation. The passive technology owners practice many energy conservation behaviors, as mentioned previously. Although the nonsolar homeowners are concerned with economic personal gains, those who are amenable to passive technologies are less concerned prefering a realistic approach to economic gains. Expected bills savings and economic preferences were not very important to them. The passive solar technology owners had the lowest scores on economic gains. Energy is perceived as a natural resource which must be conserved (Stern and Aronson, 1984). The passive energy technologies owners were negatively influenced by their participation in voluntary solar energy groups or as solar professionals (-.310). This suggests that those groups were not the best ones to use in informing the nonsolar homeowners about passive solar technologies in 1980. Since 30% (n-lOO) of the passive technology owners in the sample were professionally involved with solar energy, this indicates that energy activists seeking a soft energy path transformation should meet with such groups. Perhaps they had higher expectations of support and were discontent with 252 the amount provided for passive solar energy technologies compared to active systems' support. For example, they would know that federal policies supported active but not passive solar technology acquisition in 1980 (Axelrod, 1984). Only 12% (n-262) of the active solar technology owners were professionally involved with solar energy technologies so these groups have a smaller influence on perceptions of contextual resources as indicated by its (-.027) weight on the contextual resources variable. This is further reinforced by the finding that contextual resources is negatively, though weakly related directly to SPP in most of the solar models. LIMITATIONS AND SUGGESTIONS FOR FURTHER RESEARCH Due to the limited development of Lovins' theory of energy based social change and of the measurement problems inherent in secondary research, there are many criticisms which can be made of this study. Therefore, my focus in reviewing this work is on how this study could have been improved given the limitations of data and theory. Limitations of the Study: There are several ways in which the design of the study can be improved. The primary focus of this study was on comparing the solar and nonsolar homeowners. This is consistent with Lovins' concerns for changing the energy base of society. 253 There are two variables, income and geographic region, which influence the amount of resources available for solar energy technologies adoption for both the solar and nonsolar samples. Their influence in biasing the results must be more clearly addressed. I discussed the great differences which exist between the solar and nonsolar samples in their access and control of the income resource. The solar sample represents an elite group which controls many socioeconomic resources while the nonsolar sample represents a cross-section of all social classes. Therefore the results may be presenting the ideological orientation of a high socioeconomic class of people rather than primarily attitudes and behaviors of solar energy technology users. This bias can be mitigated by weighing the nonsolar sample to be equivalent to the income groups of the solar sample. This is most important to do for the income variable since it has such a strong influence on the contextual resources latent variable. The geographic region may also bias the nonsolar sample to underrepresent its contextual resources. I have already indicated how there are great regional variations in support for solar energy technologies through policy and physical elements. Garrett-Price et al. (1980) have indicated that in the short run, solar energy technologies will be most economically competitive in the western geographic region of 254 the U.S. As shown earlier, the solar sample overrepresents the West compared to the nonsolar sample and underrepresents the Midwest. To remove the potential regional bias, the nonsolar sample can be weighted to be more similar to the solar sample. Such regional weighting may further strengthen the differences between the solar and the nonsolar homeowners for the nonsolar homeowners will show even further support for the use of solar energy technologies than when they are randomly spread throughout the country. The bias which currently exists in the analysis due to geographic influence, tends to reduce the SPP scale scores of the nonsolar homeowners rather than increase them. However, to more accurately examine the differences in SPP the weighting should be used. Another element which is contaminating some of the responses of the solar technologies owners is the difference in the length of time which they have owned their technologies. As shown in the SPP development model, the length of ownership has a stronger influence on the active solar owners than the passive. I have argued that different historical experiences may be accounting for some of the recorded differences in the SPP scores. For example, the cohort which experienced the Depression will tend to be more 255 natural resources conserving than younger cohorts. One way to reduce the influence of history is to eliminate those who have owned the solar technologies for ten years or more from the analysis. This will also more cleanly capture the influence of the "energy crisis” decade (1970-1980). As mentioned previously, longitudinal data on the nonsolar homeowners is not available. A more refined analysis can also be done of the different solar technology types. I have argued that it was important to closely examine the active and passive solar owners as examples of two different forms of flow energy base changes. Although these data are biased to the active solar energy technology owners, there may be more passive or mixed technology users in the country. If this is true, then these owners will have greater numerical influence on societal changes. Therefore it is also important to examine the SPP scale scores of the ”mixed" solar technology owners. Further effort could also be made to identify the combination of solar energy technology owners which composed the national population in 1980. The SPP scale scores could be weighted appropriately for these data to be more generalizeable to the U.S. population of solar energy technology users of that historical period. 256 In Chapter VI, I have identified some of the measurement problems of using the data. Clearly the SPP scale must be further developed. Additional dimensions need to be added to the scale to more fully capture the notions of the soft energy path changes: (1) equity, (2) social diversity, (3) community economic development, (4) grassroots democracy, (5) community cohesion, (6) less personal alienation and (7) greater resilience (e.g. less risk). I have already discusses dimensions 1 and 2. Dimensions 3, 4, and 5 focus more strongly on using the local community as the unit of change rather than the household. The Lovinses have proposed that the community is the best unit for creating social changes as discussed in chapters II and III. Although they don't present a minimum size of population required for such changes in their written work, Hunter Lovins has verbally expressed that about 2,000 people are the minimum number necessary to make the use of conservation and renewable energy technologies a useful economic development strategy for a community (Roundtable discussion, May, 1986). The dimension of energy as a tool for local community economic development is therefore another dimension to be incorporated in the SPP scale. Grassroots democracy is the process to be used at the local 257 community level for the decision-making necessary for soft energy path social changes. Although the current study examines preferences for community self-reliance and the participation of homeowners in community activities, it does not examine preferences for grassroots democratic decision- making as opposed to policy-makers' and community leaders' decision-making processes. This dimension should include indicators of the process of decision-making as well as of control over the outcome of the decisions. As argued previously, the democractic process may simply be another method of coopting participation without power. This is an important dimension to be included in the SPP scale. The SPP scale also does not include the dimension of community cohesion. This is an outcome of the grassroots democracy process according to Lovins. Clearly this dimension should be measured with indicators of cohesion as well as of conflict since both are probable outcomes. It should also differentiate among the groups of the community who may be the winners (the populace?) and the losers (the elites?) of the social changes. The alienation dimension is presented at the individual level by Lovins. This dimension should be measured at both the individual and the community level since more or less alienation is an outcome of participation with power. It may be that the passive solar energy technology owners, for 258 example, have behaviors which are most congruent with changes to the soft energy path but are the most alienated from the local community. There are hints of this in their low scores on community self reliance and in the negative relationships of their participation in solar energy groups with the contextual resources variable. The dimension of greater resilience and less risk should have a number of indicators measuring various forms of risk. The Lovinses have proposed two major areas of greater resilience with renewable energy technologies: (1) greater energy system resilience free from widespread simultaneous "down times" as experienced with "brown outs" using the current centralized energy system: (2) greater resilience from the threat of sabotage such as that provided by terrorist control of nuclear power plants or bombs created with the by-products of power plants. Although not specifically mentioned by the Lovinses, another area which this dimension should measure is freedom from catastrophic accidents which are a potential of the centralized, nuclear power driven energy system. The inclusion of these additional dimensions in the SPP scale will strengthen its ability to measure preferences which are compatible with the development of soft energy path social changes. 259 Other latent variables of the SPP development model need better indicators. The energy vulnerability needs variable would be improved by stronger measures of perceived household energy vulnerability. Indicators of the salience of energy vulnerability and measures of actual energy use would strengthen this variable. Size of dwelling and of the household are also important to measure directly as opposed to indirectly. The technology type latent variable also needs stronger indicators. As mentioned previously, some were weak due to lack of adequate responses from the nonsolar homeowners. More indicators of participation in the operation of the technologies would strengthen this variable. So few problems were experienced by the solar technology owners that there was little variance to be interpreted by the SPP development model. Perhaps more variance will occurr when people have owned the technologies for a greater time period. This could be examined through a panel study. Finally, the model of SPP development itself should be examined in respecified form. Since there is little relationship between the use of the technologies and the development of soft path preferences, the relationship between energy vulnerability needs --> contextual resources --> SPP --> evaluation --> type of solar energy technologies 260 could be analyzed for the different solar energy technology owner samples. This would more clearly differentiate between the "hard” and "soft" streams of changing the energy base of society. Within this model, solar energy technologies are a direct expression of values. However, to examine the impacts of the implementation of these technologies on social structure and the feedback from those changes on further development of the preferences, the study must be conducted at the community level rather than at the household level as possible with these data. These are the primary criticisms and refinements suggested for this study. Obviously there are many other research questions which could also be developed. Future Research Implications: Some of the criticisms of the current study presented above can be developed into future research strategies. Obviously the development of the SPP scale requires further research. Through the suggested respecification of the model for example, a closer examination can also be made of the variables which create the positive evaluation of solar energy technologies. This research would investigate more closely the symbolic meaning of a changed energy base as 261 expressed through solar energy and other renewable technologies (e.g. wood stoves, wind, etc.). The influence of the symbolism of the change can be compared with the behaviors supportive of the ideology. Questions such as what social classes are supportive of the ideology and what social classes perform the behaviors of the ideology can be further investigated. The question of whether the population is able to withdraw from the control of the elites of society through obtaining control of energy and community processes provides a fertile ground for further research. Is it more efficient and equitable to obtain a changed energy base through the processes of policy, program and codes developed by the elites or through individual households' decisions made in marketplace competition? This research could also examine the impacts of policies, specifically the support currently provided for passive solar system acquisition. Has this reinforced the development of the soft energy path along soft lines? The conflicts between elites and the population potentially created by a change in the energy base have been inadequately examined. Is such a change really threatening to the production forces of society or are those forces simply extending their dominance over the new energy 262 resource? Is the hard energy path resisting a movement to the soft energy path or has this remained a dormant issue? The differences between energy conservation practices and other natural resources conservation and attitudes could also be pursued more closely. Are the conservers also the ones changing the energy base of society? Clearly this is true in the case of passive solar technology owners, but it does not hold for the active sample. Finally, other theoretical frameworks could be used to address questions about changing the energy base of society. Diffusion research could be used to further analyze the differences between those owning solar energy technologies for five years or more and less than that time period. I have suggested that five years is a critical ownership length. These owners could be compared to the persons who have chosen to no longer own solar energy technologies. These people may have become disenchanted with the soft energy path or found that they were not saving enough money by using the solar energy technologies. These would be important contributions to building knowledge about conservation and the dynamics of changing the energy base of society. They would provide more information which could then be built into a more systematic theory about the relationship between the energy base, technologies, and 263 social structures. This theoretical work is critically needed to provide guidance to further empirical research. CONCLUSIONS I began this study with the expectation of finding some relationships between soft path preferences with the behavior needed to change the energy base of society arguing that both the attitudes and the renewable technology change were necessary to implement the soft path social structural changes proposed by Amory Lovins. The findings indicate that there was general attitudinal support for Lovins' notions of changing the energy base in 1980. The behavioral support, though energy conservation, solar energy technology ownership, and other natural resources conservation behaviors were not as widespread. The solar, active and passive samples performed more of these than did the nonsolar sample. It is suggested that the renewable energy base implemented by the solar homeowners is proceeding along two lines, the hard energy path direction being developed by the active solar technology owners and the soft energy path line, by the passive technology owners. It is proposed that the active solar technology owners view energy as a commodity while the passive owners perceive it as a natural resource to be 264 conserved. Therefore, the active solar technology owners are continuing to reinforce the current hard energy path while the passive technology owners may be part of the vanguard of the soft energy path. This perspective is reinforced by the finding that the passive solar energy technologies influence the evaluation of the technologies and the development of attitudes and behaviors compatible with the soft energy path while the solar technologies have no influence on the evaluation and soft path preferences of the active solar technologies owners. Further research is called for to more closely examine this proposed bifurcation of soft energy path developments. ENDNOTES 1 Olsen (1983) first used this phrase to indicate individuals' preferences for policies based on renewable energy sources and conservation rather than on nonrenewable sources such as oil, coal and nuclear. 2 I will use "Lovins" to refer to the original ideas developed by Amory Lovins. The "Lovinses" is used with the more recent developments since Amory's partner, Hunter Lovins, has been jointly involved in the evolution of the SEP notion since 1980. 3 Hunter is a lawyer and has undergraduate sociology training. 4 The goal of the Fishbein model is to predict a person's behavioral intention to do something based on the person's attitudes and norms towards the behavior (Keating, 1982; Fishbein and Ajzen, 1975). 5 Schnaiberg does not present any data supporting the advocates' role in the diffusion of the technologies. The diffusion could be simply attributed to market behavior. 265 6 266 They are both small-scaled, decentralized, based on 7 local and renewable resources, and under the control of the user (Lodwick and Morrison, 1982; Lovins, 1978: Morrison and Lodwick, 1981: Morrison, 1980: Schumacher, 1973). This theme has been more strongly developed in the See The 10 In Lovinses' recent work. See especially Lovins and Lovins (1982). also various issues of the Rocky Mountain Institute (RMI) Newsletter. renewable energy technology owners often will also have nonrenewable systems as back-up while the nonrenewable owners do not have renewable system supplements (Farhar-Pilgrim and Unseld, 1982). 1984, two years after the Solar Energy and Energy Conservation Bank was put into operation under a federal district court order, bank funds were limited to conservation measures and passive solar technologies. No longer was support for active solar systems allowed (Axelrod, 1984:212). APPENDIX A: SELECTED DESCRIPTIONS OF THE SAMPLES MOVING PLANS: SOLAR AND NONSOLAR HOMEOWNERS 267 Table 24 Solar Nonsolar % (n) % (n) MOVING PLANS Very unlikely 45 (1672) 57 (1157) Unlikely 28 (1042) 20 (397) Unsure 11 (425) 7 (142) Likely 5 (189) 6 (124) Very likely 11 (414) 10 (197) Missing data 2 (67) 0 (6) Total 3809 2023 268 PERCEPTIONS OF THE Table 25 "ENERGY CRISIS" Nonsolar Solar % (n) % (n) LIFE STYLE IMPACTS: Serious changes 9 (179) 5 (205) Less comfort 30 (610) 27 (1032) Some effects .5 (10) 4 (148) No effects 47 (942) 52 (1975) Don't know 14 (279) 12 (449) Missing Data .1 (3) -- ---- Total 100 (2023) 100 (3809) i - 3.263 X'- 3.376 s.d. - 1.268 s.d. - 1.156 U.S. ENERGY FUTURE: Worst: 1 0 (2) .l (3) 2 .2 (5) .1 (6) 3 3 (57) l (48) 4 16 (298) 13 (474) 5 8 (140) 9 (322) Same: 0 15 (305) 12 (424) 6 13 (267) 19 (674) 7 33 (621) 38 (1355) 8 8 (156) 8 (274) 9 .5 (10) .6 (21) Better: 10 0 (l) .1 (4) Missing data: 8 (161) 5 (204) Total (2023) (3809) 269 Assay v Amosmv mm Ammmv ma smegma ozone nmfiom Amoev m Aspens mm Ammov es soccemmmeouo madam sumo msflmmez oz mm> “menses xuosums nmaom Emcee m lose m Amose e “nose m Lemme as Asmmsv me masomcoz Aeemv A Away A Lame s Remy m Ammme A Amssme mm paeom “mcoflcflmo msoonmflmc\mocmflum mo mosmuuomsH sumo ucuumsH ucmuuomsH .usuumsH .usuumsH.ucuuomaH .mmez sum> parsmeom saurmssm uoz Assoc Hm Anna. cs Ammee on Amemv om uaaomcoz Loses as 154V H Aemssv mm Ammmsv he madam muonnmflmC\mosmfium Leone on Lame m Ammme mm Lemme we unsomcoz Amuse ms Lens H Ammvse he lessee mm masom MOHMflOOmmm no.0 Aeemv em assay as Amemv ms Aneoae as nasomcoz «None 8 Lame H Amsme G Amemne mm uaaom mmsomm mcflmmfiz ommommo omxwz manmuo>mm Bum mom BmommDm MKOBBMZ Qfizommmm "mmumDOmmm A4 Annoy em Amuse m Acme m Ammv m unsomcoz “meme 8 Assay m Lemme Ha Lmflee NH Ammmv ms Amqosv 6m unsom OOMHO>OO OOCOHSMGH Loose m finesse om Lemme on Aooae m Acme m knee m unsomcoz Lemme A Aomose on Amway em Anmee ms Amway m Amery mm madam momsm>oo >ucsuum3 Ansel s Amway ms Amoee Hm Refine ms Asonc 6H Amuse mm uaaomcoz Loewe A “wee m Amssv m Aemsv s Amemv m Asommv mm uaaom >cmmsoo wuflaeub "mzmqmomm quezmeom sumo ucuuQEH ucmuuomsH .ucuuQEH .ucuumsH.ucuuomsH .mms: sum> oarsmeom saunmeHm uoz hm OHQMB emommbm Q Pumas: w> wammoz pagans: o> ...... wmom o> ..u «mo.— .mm_: a_mcotpm \ogsmca aFucoium mze_t<=zu==¢um¢ ez< mac—pzmh21 "ze_h<=s<>m mm space SPP: Table 34 277 SELF RELIANCE PREFERENCES very Somewhat Slightly Not Missng Imprtnt. Important Imprtnt. Imprtnt Imprtnt. Data General self reliance Solar 38 (1351) 31 (1077) 15 (527) 8 (275) 9 (299) 7 (280) Nonsolar 34 (640) 39 (726) 15 (272) 7 (138) 5 (92) 8 (155) ENERGY SELF RELIANCE Fewer big power plants Solar 16 (550) 17 (588) 20 (689) 20 (685) 28 (980) 8 (317) Nonsolar 31 (587) 38 (710) 16 (296) 8 (149) 7 (139) 6 (142) Independence of utilities Solar 26 (916) 23 (807) 17 (582) 14 (479) 21 (749) 7 (276) Nonsolar 39 (747) 30 (566) 14 (272) 10 (183) 7 (134) 6 (121) Reliable energy supply Solar 23 (817) 26 (897) 19 (658) 14 (486) 19 (651) 8 (300) Nonsolar 45 (860) 38 (720) 9 (168) 5 (85) 4 (69) 6 (121) LOCAL SELF RELIANCE Inc. self reliance skills* Solar 50 (1867) 12 (444) 14 (534) 19 (708) 4 (152) 3 (104) Nonsolar 42 (849) 12 (235) 15 (299) 18 (360) 13 (270) 1 (10) Community resolutions Solar 34 (1282) 52 (1962) 8 (295) 4 (156) 1 (51) 2 (63) Nonsolar 19 (378) 66 (1336) 10 (199) 5 (98) 1 (10) 0 (2) Independence from national policies Solar 16 (542) 12 (421) 12 (415) 13(445) 47(1636) 9 (350) Nonsolar 38 (700) 31 (569) 14 (252) 9(172) 8 (149) 9 (181) Not federal resolution *9 Solar 29 (1084) 33 (1239) 16 (606) 15(542) 7 (275) 2 (63) Nonsolar 11 (224) 32 (641) 20 (395) 28(574) 9 (187) o (2) 8 Translated from 5-point frequency of behavior responses ** Translated from 5-point "agreement“ responses 278 Table 35 SPP: RESOURCES CONSERVATION PREFERENCES very Somewhat Slightly Not Missng Imprtnt. Tmprtant Imprtnt. Imprtant Imprtnt. Data General conservation Solar 28 (990) 28 (986) 19 (674) 13 (464) 12 (408) 8 (287) Nonsolar 44 (850) 36 (694) 11 (218) 5 (100) 3 (53) 5 (108) Ease energy shortage Solar 26 (925) 30 (1064) 22 (768) 13 (468) 9 (333) 7 (251) Nonsolar 34 (662) 39 (749) 15 (285) 8 (151) 5 (89) 5 (87) Decrease personal use at Solar 50 (1861) 43 (1627) 3 (111) 3 (112) 1 (44) 1 (54) Nonsolar 29 (594) 58 (1162) 5 (105) 7 (134) l (23) o (5) CONSERVATION BEHAVIORS * Recycle home materials Solar 14 (537) 22 (831) 13 (470) 35 (1292) 16 (592) 2 (87) Nonsolar 12 (236) 11 (212) 9 (179) 24 (478) 45 (908) 1 (10) Recycle clothes Solar 1 (22) 2 (70) 5 (194) 36 (1335) 56(2080) 3 (108) Nonsolar 1 (22) 2 (42) 6 (117) 33 (661) 58(1164) 1 (17) Contribute to ecological organization Solar 14 (527) 7 (268) 25 (934) 6 (213) 47(1734) 4 (133) Nonsolar 3 (51) 3 (65) 13 (263) 3 (64) 78(1552) l (28) * Translated from S-point frequency responses ** Translated from 5-point "agreement“ responses 279 Table 36 SPP: ENERGY CONSERVATION ACTIVITIES SOLAR NONSOLAR Performed na* mv** Performed na* mv** % (n) % % % (n) % % Thermostat at 65 in winter 13 (1842) 25 7 64 (1284) 7 1 Thermostat at 78 in summer 34 (1304) 48 8 35 (716) 36 3 Carpoo1 26 (964) 26 12 22 (436) 30 3 Small car owner 65 (2475) 7 9 45 (919) 9 2 Window covers 52 (1995) 14 10 48 (979) 5 3 Caulking 61 (2306) 18 9 76 (1539) 3 2 Storm windows 53 (2005) 21 10 60 (1204) 7 2 Hot water temp. control 63 (2406) ll 10 59 (1189) 2 3 Fireplace device 32 (1207) 36 ll 22 (434) 38 4 Efficient appliances 52 (1949) 12 10 37 (739) 8 5 Insulated 54 (2084) 19 10 61 (1242) 4 2 Heat pump 11 (409) 31 3 4 (78) 12 5 Furnace timer 16 (612) 30 12 12 (248) 4 10 Wood stove 30 (1148) 23 11 18 (356) 7 3 na* - not applicable mv** - missing data 280 Table 37 SPP: PERSONAL GAINS PREFERENCES very Somewhat Slightly Not Missng Imprtnt. Imprtant Imprtnt. Imprtant Imprtnt. Data % (n) % (n) 8 (n) 2 (n) % (n) % (n) ECONOMIC Rising cost protection Solar 57 (2038) 30 (1090) 7 (256) 3 (94) 3 (123) 6 (208) Nonsolar 48 (919) 38 (731) 9 (165) 3 (60) 3 (48) 5 (100) Decreased utility bills Solar 47 (1674) 29 (1044) 12 (443) 7 (240) 6 (197) 6 (211) Nonsolar 57 (1090) 30 (568) 8 (144) 3 (60) 3 (56) 5 (105) Increased home resale value Solar 8 (281) 16 (567) 20 (695) 21 (749) 35(1252) 7 (265) Nonsolar 30 (568) 35 (665) 16 (300) 11 (215) 8 (160) 6 (115) NONECONOMIC Increased home comfort Solar 14 (479) 20 (701) 17 (594) 13 (459) 36(1266) 8 (310) Nonsolar 32 (591) 43 (794) 13 (231) 7 (123) 6 (112) 9 (172) Increased personal prestige Solar 4 (133) 6 (211) 10 (335) 13 (461) 67(2356) 8 (313) Nonsolar 5 (102) 10 (185) 10 (182) 13 (253) 62(1183) 6 (118) (Nonsolar only) Good investment* 11 (230) 22 (445) 33 (670) 22 (443) 11 (223) 1 (12) Long term money savings Solar 51 (1843) 30 (1095) 10 (352) 5 (177) 4 (139) 5 (203) Nonsolar 56 (1067) 32 (612) 7 (137) 2 (43) 3 (50) 6 (114) * Translated from 5-point money saved assessment APPENDIX B: OPERATIONALIZING THE SPP MODEL APPENDIX B OPERATIONALIZING THE SPP MODEL SOLAR: NONSOLAR: ENERGY VULNERABILITY NEEDS A. Needs created by family and housing structures: XHOUSE HOUSE What type of structure do you live in? [1] - one family house [2] - attached house (duplex,townhouse,etc.) [3] - mobile home [4] - other [5] - no response *XAVCOSTV - *AVCOSTV Average monthly energy cost (XWNCOST + XSUMCOST / 2)/1o same [5000-0000] XWNCOST WINCOST During this past winter - say during December, January, and February - what were your average monthly energy costs for electricity, heating, and hot water? [0000-8000] - number of dollars [0000-1300] XSUMCOST SUMCOST During this past summer - say June, July, and August - what were your average monthly energy costs for electricity, heating, hot water, and cooling? [0000-6000] - number of dollars [000-900] XNUMCHIL NUMCHIL number of children [20-00] - number of children same XFAMTYP FAMTYPE household structure [1] - two parent family [2] extended family (e.g.multigenerational) same [3] married couple [4] single parent [5] one adult 281 282 SOLAR NONSOLAR XRAGE RAGE Resource ordered age of respondent: [1] - 35 - 44 [2] - 45 - 54 [3] - 55 - 64 [4] - 25 - 34 [5] - 65 and older [6] - 24 or less B. Needs related to perceptions of the ”energy crisis": *XLADDRV I LADDRV = Expected energy situation in our country in five years. XLADDRl - XLADDR2 . LADDRl - LADDR2 [0-10] - estimate of whether the energy situation will be better or worst XLADDRl LADDRl Here is a picture of a ladder. Let's suppose the top of the ladder represents the best possible energy situation for our country and the bottom, the worst possible energy situation for our country. Please show me on which step of the ladder you think the United States is at the present time. XLADDR2 LADDR2 Just as your best guess, if things go pretty much as you now expect, where do you think the U.S. will be on the ladder, let us say, about five years from now? [0 - 10] - 0 - worst 10 - best XIMPACT IMPACT In different parts of the nation, people have reported a variety of effects of the same energy situation on their daily lives. How has the energy situation affected your family (household)? Would you say...? [1] - We have had to make serious changes in our daily habits. [2] - Our life has been less comfortable and convenient, but it is not serious. [3] - Don't know. [4] - We have had to make a few adjustments, but our lifestyle has not been affected. [5] - It really has had no effect on us. 283 SOLAR NONSOLAR RESOURCES A. Personal Support: XEDUC REDUC What is the highest level of education you have completed? [11 graduate work or more [2] college graduate same [3] some college [4] - trade or technical school [5] high school graduate [6] - less than high school [9] $7000 - 9999 a year [10] under $5000 - 6999 a year *XPERSUP = *PERSUPV a Perceived personal network support XINCOME INCOME Please check off which category best Which category best represents the total annual income, represents the before taxes, of your immediate family total annual income living in your household? of all persons [1] - over $55,000 a year living in your [2] - $45,000 - 54,999 a year household? [3] - $35,000 - 44,999 a year [4] - $25,000 - 34,999 a year same [5] - $20,000 - 24,999 a year [6] - $15,000 - 19,999 a year [7] - $12,000 - 14,999 a year [8] - $10,000 - 11,999 a year (XFOSPOU + FOFRND*2 + FOWORK*3) same [1] - support [2] - neutral [3] - no support Please indicate whether each of the following is favorable to, or opposed to, using solar energy for homes. [1] - favorable [2] - unsure [3] - opposed XFOSPOU FOSPOUS Your spouse/housemate XFOFRND FOFRNDS Your neighbors and friends XFOWORK FOWORK People you work with 284 SOLAR NONSOLAR B. Institutional Support: *XNSTSUP *INSTSUP Perceived institutional support - (XCRUTIL + XCRWARR + XCRINSU + XCRFIRM) same [1] - great support [2] - support [3] - some support same [4] - little support [5] - no support ..which of these were concerns to you ..how important a when you were thinking about using solar concern it would energy in your home... show how important be for you, if that concern was to you. you were thinking about using solar [1] - not at all important energy in your [2] - slightly important home. [3] - somewhat important same [4] - important [5] very important XCRUTIL CRUTIL Possible problems with utility company XCRWARR CRWARR Warranty coverage for solar energy system XCRINSU CRINSU Cost and difficulty of getting solar energy systems covered by homeowners insurance XCRFIRM CRFIRM Problems with dependability of firm C. Governmental Support: *XGOVSUPV - *GOVSUPV - Perceived governmental support for solar energy technologies. (XFOFED*3 + XFOSTAT*2 + XFOLOCAL) same 285 SOLAR NONSOLAR Please indicate whether each of the following is favorable to, or opposed to, using solar energy for homes. [1] - favorable [2] - unsure [3] - opposed XFOFED The federal government XFOSTAT Your state government XFOLOCAL Your local government *XFEDSUP - Received federal support (XTAXFED + XHUDl + XHUD2) [l] - great support [2] - support [3] - some support [4] - no support Types of federal support received: XTAXFED Federal tax credit [1] - applies [0] - does not apply XHUDl HUD Solar Heating and Cooling Demonstration Program Grant XHUD2 HUD Solar Hot Water Demonstration Program *XSTASUP - Received state support (XTAXST + XTAXPRO + XLOAN) [1] - great support [2] - support [3] - some support - no support FOFED FOSTATE FTAXINC Do you happen to know if there is a federal credit available on your income tax for those who pur- chase solar energy systems for the home? [11 ‘ Yes [2] unsure [3] - no 286 SOLAR Type of state support received: [1] - applies [0] - does not apply XTAXST State Sales tax credit or exemption XTAXPRO State Property tax exemption XLOAN State low income loan *XLCLSUP - Received local support (XTAXLCL + XTAXPLC) [1] - support [2] - some support [3] - no support XTAXLCL Local Sales tax credit or exemption [l] - applies [0] - does not apply XTAXPLC Local Property tax exemption *XPRCOSTV - Problems with increased costs (XPRFIN + XPRUTIL + XPRTAX + XPRVALU) [0] - no problems [1] - slight problems [2] - some problems - problems - many problems As a solar owner, which of the following problems have you actually experienced? [0] - Does not apply [1] - Applies NONSOLAR STAXINC As far as you know does your state have an income tax credit for the purchase of solar energy systems for the home? [1] Yes [2] unsure [3] no 287 SOLAR NONSOLAR XPRFIN --- Difficulty in obtaining financing XPRUTIL --- Utility raised rates for solar owners XPRTAX --- Increased property tax XPRVALU --- Resale value of house has decreased *XPRLCLSV I Local policy problems (XPRSAR + XPRCODE) [0] - no problems --- [1] - some problems [2] - several problems As a solar owner, which of the following problems have you actually experienced? [1] - does not apply [0] - applies XPRSAR Difficulty obtaining solar access rights --- XPRCODE Conflicts with local building codes --- *XPRPEOPVI --- Problems from people (XPRVAND +’XPRNEBR + XPROBST) [0] no problems [1] - some problems [2] - problems - many problems As a solar owner, which of the following problems have you actually experienced? [0] - does not apply [1] - applies 288 SOLAR NONSOLAR XPRVAND --- Vandalism to the solar energy system XPRNEBR --- Opposition of the neighbors XPROBST --- Obstruction of your solar energy system by vegetation, new buildings, etc. *XPRBLDV I Problems acquiring the solar system --- (XPRINST + XPREVAL + XPRINFO + XPRQUAL) [0] - no problems [1] - slight problems [2] - some problems [3] - problems [4] - many problems As a solar owner, which of the following problems have you actually experienced? --- [0] - does not apply [1] - applies XPRINST Installation problems --- XPREVAL Difficulty in evaluating and choosing --- among competing systems XPRINFO Lack of clear, reliable information --- XPRQUAL Finding a qualified contractor, --- architect, or builder *XPRFRNDV I Experience with personal support (XMEMBER + XPROINV) --- [l] - great support [2] - support [3] - little support 289 XMEMBER --- Are you a member of a solar energy groups, such as a local solar energy association or a chapter of the International Solar Energy Society? [1] - yes [0] - no XPROINV --- Are you professionally involved with solar energy? [1] ’ YES [0] - no *********************************************************** ACTIVE SYSTEMS (solar only) PASSIVE SYSTEMS TYPE OF SOLAR TECHNOLOGY XOWNTIMV same Number of years the solar technology is owned [000-652] - number of months/12 XDATINlV XPYRlV Year active system 1 was installed in 10 Year passive system 1 [l - 80] - year/10 was constructed in 10 [1 - 80] - year/10 XCOSTSEV same What was the total installed cost of your energy system(s)? [0000-9996] - number of dollars/1000 *XPROPTNVI same Problems operating the solar technologies (XPRMAIN + XPRPART + XPRHEAT + XPRDESN) [0] no problems [1] slight problems [2] - some problems [3] problems [4] - many problems As a solar owner, which of the following problems have you actually experienced? [0] - Does not apply [1] - Applies XPRMAIN --- Maintenance problems 290 ACTIVE SYSTEMS (soLar only) XPRPART Difficulty finding parts and equipment for initial purchase or repair XPRHEAT Overheating XPRDESN Difficulties arising from improper system design *XABLDMTD Method used in building active system (XABLDI * .5) (XABLD3 * 1.5) + (XABLDZ) + XABLD1 - sum of values I l XABLDZ - sum of values I 2 XABLD3 - sum of values I 3 (XAMANUl + XAMANUZ + XACBUILl + XACBUIL2 + XAINSTl + AINSTZ) XAMANUl Active System 1 was manufactured by [1] - small local firm + myself (volunteered) [2] - national firm [3] - foreign firm (+ other) XAMANUZ Active System 2 was manufactured by XACBUILl Active System 1 built [1] - systems built on site [2] - systems bought as components and assembled on site [3] - systems bought as a complete "packaged“ system and assembled on site (+ other) XACBUILZ Active System 2 built PASSIVE SYSTEMS *XPBLDMTD Method used in building passive system same - with XPBLDl, 2, 3. same with passive variables same with passive variables XPDESIl Passive System 1 designed by [1] - yourself [2] - acquired plans [3] - a contractor, an architect, other XPDESIZ Passive System 2 designed by XPBUILl Passive System 1 built XPBUIL2 Passive System 2 built 291 ACTIVE SYSTEMS (solar only) PASSIVE SYSTEMS XAINSTl Active System 1 installed XPINSTl [1] - by yourself Passive System 1 [2] - by both self and contractor installed by [3] - by a contractor (+other) same XAINST2 XPINST2 Active System 2 installed by Passive System 2 installed by From whom did you purchase your system(s) or parts and materials for it? [1] - applies [0] - does not apply *XLCLAPRT *XLCLPPRT Sources of parts and materials for either of the solar systems: [1] - XASOCOl or XASOCOZ XPSOCOl or XPSOCOZ Local solar company [2] - XAHARD1 or XAHARDZ XPHARD1 or XPHARDZ Local hardware, lumber, etc. supply house [3] - XAHVACIl or XAHVACIZ XPHVACl or XPHVACZ Local heating, plumbing, ventilating, air conditioning store [4] - XABLDRl or XABLDR2 XPBLDRl or XPBLDR2 Local building contractor [5] - XAARCHIl or XAARCHIZ XPARCHl or XPARCHZ Local architect [6] - XACHAIl or XACHAIZ XPCHAIl or XPCHAIZ Local outlet of chain store [7] - XAMAILl or XAMAIL2 XPMAILl or XPMAIL2 Mail order from a non-local solar manufacturer [8] - XANLSCl or XANLSC2 XPNLSCl or XPNLSCZ Non-local solar company *XADOWNV I *XPDOWNV I Average active systems incidents of Average passive failure incidents of failure (XADOWNl + XADOWN2)/2 (XPDOWNl + XPDOWN2)/2 XADOWNl XPDOWNl Active system 1 incidents of failure Passive system 1 [000-995] - number of incidents same XADOWN2 XPDOWN2 Active system 2 incidents of failure Passive system 2 same 292 SOLAR NONSOLAR EVALUATION: *XGENEVAL I *GENEVAL I General evaluation of solar systems (XOVEVAL + XSATEXP) (SPEVAL + OVEVAL) XSATEXP SPEVAL Altogether, how satisfied are you with Given what you know your experiences as a solar owner? about solar energy [1] - very satisfied right now, do you [2] - satisfied strongly favor, [3] - unsure, no opinion yet favor, oppose or [4] - dissatisfied strongly oppose the [5] - very dissatisfied idea of using it in your home? [1] - strongly favor [2] - favor [3] - unsure [4] - oppose [5] - strongly oppose XOVEVAL OVEVAL Based on your understanding of solar energy for homes, how do you feel about it? [1] - strongly favor [2] - favor [3] - unsure/neutral same [4] - oppose [5] - strongly oppose *XTECEVAL I *TECEVAL I Specific evaluation of technological capabilities (XBEFFl + XBEFF2 + XBEFF3 + XCRCLIM + same XCRSAFE + XCRRELI + XCROBS) XBEFFl BEFFl In general, solar energy systems are technically and economically practical today for homes. [1] - strongly agree [2] agree [3] unsure same [4] disagree [5] strongly disagree 293 SOLAR NONSOLAR XBEFFZ BEFFZ If they are practical at all, solar energy systems are only practical for homes in warm, sunny climates (reversed order of responses) XBEFF3 BEFF3 Solar energy systems in general will perform satisfactorily over a long period of time. XCROBS CROBS Possibility that solar energy systems now on the market will soon be obsolete and better systems will be available later. ..which of these were concerns to you ..how important a when you were thinking about using solar concern it would energy in your home... show how important be for you, if that concern was to you. you were thinking about using solar [1] I not at all important energy in your [2] - slightly important home. [3] I somewhat important same [4] I important [5] I very important XCRSAFE CRSAFE Safety of solar energy systems XCRRELI CRRELI Operating reliability of solar energy system XCRCLIM CRCLIM Climate too cold or cloudy or not enough wind. *XXSAVl I Difference between expected savings and actual savings (xsxpsav - XSAVING) 294 SOLAR NONSOLAR XEXPSAV II- When you bought or built your solar energy system(s), approximately what percent of your total monthly utility cost did you expect to save? [1] I over 85% [2] - 75% [3] 65% [4] 55% [5] 45% [6] 35% [7] 25% [8] 15% [9] - 5% [10]- Did not expect to save at all XSAVING III Since you acquired your solar energy system, what percent of your total monthly utility costs do you think you have actually saved? [1] I over 85% [2] 75% [3] 65% [4] 55% [5] 45% [6] 35% [7] 25% [8] 15% [9] 5% [10] Did not expect to save at all --- INVEST . At present costs, do you think that owning a solar energy system will result in sufficient utility savings to make it a good investment? [1] I definitely [2] I I think so [3] I don't know [4] I probably not [5] I definitely not 295 SOLAR XADSOL How likely is it that you will add other solar applications to your present home? [1] I very likely [2] I likely [3] I unsure [4] I unlikely [5] I very unlikely XRECMEN On the basis of your experience, have you reccommended solar energy to others or not? [1] I yes have reccommended it highly [2] yes, with some reservations [3] have not reccommended for or against it [4] I no, have reccommended against it NONSOLAR BI To what extent, if any, have you considered invest- ing in a solar system for your house in the next 2-3 years? [1] I plan to invest [2] may invest [3] don't know [4] will not invest [51 not considered EASE ...would you say your house could easily use solar energy...or could not use it? [1] I could use easily [2] I could use with some difficulty [3] I don't know [4] I ...only with major renovation [5] I could not use 296 SOLAR NONSOLAR III WILLPAY How much more than your present total monthly bill for utilities would you be willing to pay per month for your own solar energy system in your home? [1] I more than $85 [2] I $85 per month [3] I $40/month [4] I $25/month [5] I $8/month [6] I no more [7] I don't want *XTECSUPV I Evaluation of technical support (XWARSAT + XREPAIR + XINHELP) XINHELP How helpful were the [sic. operation and maintenance] instructions to you? [1] I extremely helpful [2] I helpful [3] don't know III [4] not helpful [5] harmful XWARSAT How satisfied are you with your warranty coverage? [1] I very satisfied [2] satisfied [3] don't know [4] dissatisfied [5] very dissatisfied XREPAIR Relative to the repair service for conventional heating or hot water systems, how would you rate the quality of your solar repair service? [1] I never needed repairs [2] I better [3] I about the same [4] I less reliable but adequate [5] I inadequate 297 ACTIVE SYSTEMS (solar only) PASSIVE SYSTEMS 311326.355?""""""""“""""""""mIEESEG'J" Overall condition of both systems (xaconol + XACOND2)/2 (XPCONOl + XPCON02)/2 £26133; """""""""""""""""""""" £92851 """" What overall condition describes your (active) solar energy systems 1 and 2 right now? same [1] excellent condition, never were any problems [2] very good, the few problems have been solved [3] good, some minor problems remain [4] fair, system problems difficult to resolve [5] poor, system has chronic problems XACOND2 XPCONO2 SOLAR NONSOLAR e*****s***e***e**seeens*e*seessss4*e***ss******s*s*s******* SOFT PATH PREFERENCES *XEYNDPV I *EYNDPVI Energy self reliance same (XRASELF + XRARELI + XRAUTIL + XRAPLNT) (EYSUP + UTLIND + EAPLNTS + EASREL) ...list of factors that might have ...advantages that influenced your decision to adopt could enter into solar energy...how important [was] your decision this advantage to you in making your about using solar decision to use solar energy. energy in your home. [1] I very important ...how important it [2] I important would be for you in [3] I somewhat important making such a [4] I slightly important decision [5] I not at all important XRASELF EASREL Increasing overall self-reliance XRARELI EYSUP Having a more reliable supply of energy XRAUTIL UTLIND Increasing independence from utility company 298 SOLAR XRAPLNT Reducing need for more large power plants *XLCLNDPV I Local self reliance (XLOCAL + XTRUST + xnarsos + XSKILLS) ...list of factors that might have influenced your decision to adopt solar energy...how important [was] this advantage to you in making your decision to use solar energy. [1] I very important [2] important [3] somewhat important [4] slightly important [5] not at all important XLOCAL We should pay more attention to the particular energy needs of each city or town I and to meeting those needs through local resouces wherever possible. [1] I strongly agree [2] agree [3] unsure [4] disagree [5] strongly disagree XTRUST People like me should trust the federal government to find a solution to the energy crisis. [1] I strongly disagree [2] I disagree [3] I unsure [4] I agree [5] I strongly agree XRAFEDS Increasing independence from federal government policies NONSOLAR EAPLNTS (LOCAL + TRUST + GOVIND + SKILLS) ...advantages that could enter into your decision about using solar energy in your home. ...how important it would be for you in making such a decision LOCAL same TRUST same GOVIND 299 SOLAR NONSOLAR Please indicate whether and how much you or members of your household engage in these activities: XSKILLS Developing and using skills to increase self-reliance, such as in carpentry, car repair, food preservation. [1] I regularly [2] I usually [3] I frequently I sometimes I never *XCNSBLFV I Resource conservation beliefs (XCUTCON + XRASHOR + XRACONS) ...list of factors that might have influenced your decision to adopt solar energy...how important [was] this advantage to you in making your decision to use solar energy. [1] I very important [2] I important [3] I somewhat important [4] I slightly important [5] I not at all important XRASHOR Easing the energy shortage XRACONS Conserving natural resources: protecting the environment XCUTCON People like me have a responsibility to help resolve our country's energy problems by cutting back on consumption, even if this means making some sacrifices in the way I live. [1] I strongly agree [2] agree [3] unsure disagree strongly disagree H b u-a llll SKILLS *CNSBLFV =- (CUTCON + EASHORT + EASRES) ...advantages that could enter into your decision about using solar energy in your home. ...how important it would be for you in making such a decision EASHORT EASRES CUTCON same SOLAR NONSOLAR *XCNSBHRV I *CNSBHRV Conservation behavior (RECYCL + ECOORG + (XRECYCL + XECOORG + XBUYCLO) BUYCLO) Please indicate whether and how much you or members of your household engage in these activities: XRECYCL RECYCL Recycle the newspapers, glass or cans used at home. [1] recycle all this material [2] recycle most of this material same [3] about half of this material [4] I some of this material [5] I never recycle XECOORG ECOORG Contribute to ecologically oriented organizations (such as the Sierra Club, etc.) [1] I contribute regularly to 2 or more organizations I contribute regularly to 1 organization same [3] I occasionally contribute [4] I used to contribute but no longer do [5] I never have contributed XBUYCLO BUYCLO Buying clothing at a garage sale or a second-hand store. [1] all of the household clothing [2] most items [3] I about half of the household's clothes same [4] I a few items [5] I none of the household clothing *XECCONSV I *ECCONSV I number of energy conservation behaviors which are performed by the household same [XECTH65 + XECTH78 + XECPOOL + XECCAR + XECSHUT + XECCALK + XECWIND + XECHWH + same XECFIRE + XECAPPL + XECINSU + XECPUMP + XECTIMR + XECSTOV] XECTH65 ECTH65 Set thermostat at 65 or lower in winter XECTH78 ECTH78 Set thermostat at 78 or higher in summer 301 SOLAR NONSOLAR XECPOOL ECPOOL Carpool XECCAR ECCAR Drive a small car XECSHUT ECSHUT Install shutters or shades on windows XECCALK ECCAULK Install weatherstripping or caulk around doors or windows XECWIND ECWINDO Install storm or doublepane windows XECHWH ECHWH Install temperature setting on hot water heater XECFIRE ECFIRE Install glass doors, heatilator, or other energy saving device on fireplace XECAPPL ECAPPL Buy energy-efficient appliances XECINSU ECINSU Add insulation XECPUMP ECPUM Install a heat pump XECTIMR ECTIMER Install timer on heating system or furnace XECSTOVE ECSTOVE Install wood stove *XCMFRTV I *COMFRTV = SET as comfort and prestige (XRACOMF + XRAPRES) (CMFRT + STATUS) XRACOMF CMFRT Increasing comfort of home XRAPRES STATUS Increasing status, prestige, and self-esteem *XINVESTV I SET as financial investment (XRACOST + XRABILL + XRASAVE + XRAVALU) ...list of factors that might have influenced your decision to adopt solar energy...how important [was] this advantage to you in making your decision to use solar energy. [1] I very important [2] I important [3] I somewhat important [4] I slightly important [5] I not at all important XRACOST Protecting against rising energy costs XRABILL Reducing utility bills now XRASAVE Saving money over the long term XRAVALU Increasing resale value of home NONSOLAR (EACOSTS + UTLBLS + SAVMNY + RESALE) ...advantages that could enter into your decision about using solar energy in your home. ...how important it would be for you in making such a decision EACOSTS UTLBLS APPENDIX C: MODELS' CHARACTERISTICS 303 Table 38 PEARSON CORRELATIONS: SOLAR HOMEOWNERS Variable 1 4 5 1.FAMTYPE 2.NUMCHIL .542 3.AGE .396 .392 4.LADDER .096 .060 .106 5.HOUSE .097 .068 .078 .049 6.IMPACT .068 .060 .077 .064 .024 7.AVCOST .059 .097 .050 .007 -.004 .049 8.LOCAL .070 .056 .072 .011 .029 .095 .057 9.STATE .114 .068 .059 .023 .028 .076 .060 .517 10.FEDERAL .199 .139 .186 .025 .080 .097 .104 .352 .299 11.EDUCATION .060 .008 .088 .142 .024 .072 .041 .059 .064 12.INCOME .265 .166 .283 .085 .095 -.039 .198 .056 .049 13.PERSUP .023 -.005 .015 .002 .007 .011 -.022 -.014 .007 14.GRMEMBR .030 -.010 .019 .092 .004 .026 -.031 .010 .031 15.0WNLENGTH -.191 -.131 .176 -.022 -.062 -.056 -.022 -.108 .123 16.PASSYR -.O76 -.O39 .062 .018 -.004 .019 .200 .069 .020 17.ACTYR -.117 -.107 .096 .005 -.042 -.046 -.069 -.110 .098 18.COSTTECH -.036 -.074 .094 -.054 -.010 .015 -.035 .004 .003 19.ACTBLDMTD -.021 -.017 .027 .030 -.028 .001 I.077 -.044 .034 20.PASBLDMTD .078 .066 .031 -.032 .020 .015 .037 .002 .019 21.PRCOST -.041 -.056 .008 -.054 .017 -.066 .017 -.028 .036 22.PRLCL -.037 -.027 .033 -.067 .019 -.084 .031 -.072 .054 23.PRPEOP .005 -.012 .023 -.068 .044 -.048 -.019 -.084 .074 24.PRBLD -.050 -.047 .073 -.088 .004 -.087 -.029 -.129 .102 25.PROPNG -.005 -.010 .027 -.115 .004 -.081 -.017 -.016 .005 26.GENEVAL .004 -.012 .006 -.030 .012 -.018 -.103 .013 .026 27.RECMEN .024 .008 .015 -.065 .031 -.004 -.096 .017 .029 28.ACTCOND .055 .009 .027 .035 .033 -.062 -.019 .048 .038 29.PASSCOND .125 .061 .093 .126 .085 .018 -.028 .091 .091 30.TECEVAL -.069 -.069 .084 .026 -.002 -.098 -.060 -.096 .061 31.ADDTECH .071 .046 .082 .026 .028 .097 .020 .069 .085 32.TECHSUP .009 .000 .014 -.033 .019 .019 .024 -.011 .011 33.EXPSAVE .015 .029 .010 -.001 -.016 -.001 .053 -.013 .030 34.EYSLFREL .022 .005 .048 -.014 -.005 .071 -.066 .056 .057 35.CNRBLF -.019 -.021 .005 -.057 -.003 .050 -.030 .048 .035 36.STATUS I.013 -.051 .013 -.018 -.022 .013 -.013 .025 .021 37.LCLSLFREL .040 -.005 .049 .079 .036 .070 -.065 .035 .038 38.ECON .096 .099 .119 -.059 .011 .056 -.043 .040 .042 39.CNRBHVR -.037 -.040 .037 .040 -.007 .049 -.042 .065 .070 40.EYCONS .053 .073 .031 .054 -.011 .073 .001 .001 .045 304 Table 38 (cont.) 10 11 12 13 14 15 16 17 18 11.EDUCATION .095 12.INCOME .212 .310 13.PERSUP -.007 .062 .018 14.GRMEMBR -.049 .140 -.030 .066 15.0WNLENGTH -.304 -.065 -.156 .021 .088 16.PASSYR -.028 .012 .045 .035 -.008 .445 17.ACTYR -.261 -.012 -.156 .039 .050 .129 -.017 18.COSTTECH .007 -.087 -.166 .002 -.125 .071 .133 -.027 19.ACTBLDMTD -.101 -.070 -.157 .000 .029 .063 .004 .127 .087 20.PASBLDMTD .056 -.145 -.017 -.041 -.149 -.035 -.014 -.260 .268 21.PRCOST .006 .006 .056 .009 -.110 -.004 .065 -.029 .086 22.PRLCL -.013 -.047 .007 .019 -.130 .009 .072 -.015 .093 23.PRPEOP -.033 -.047 .016 .019 -.104 -.037 -.063 .013 .006 24.PRBLD -.091 -.123 -.055 .049 -.024 .057 .074 .059 .128 25.PROPNG .094 -.126 .020 .025 -.122 -.066 .031 .010 .166 26.GENEVAL .043 .078 -.044 .139 .124 .051 .030 .035 .073 27.RECMEN .042 .046 -.050 .107 .107 .059 .022 .052 .084 28.ACTCOND .144 -.012 .058 .047 -.002 -.038 .060 -.017 .106 29.PASSCOND .098 .274 .069 .076 .171 .015 .049 -.399 -.137 30.TECEVAL -.108 .084 -.031 .096 .130 .130 .038 .072 .044 31.ADDTECH .006 .067 -.054 .065 .241 .029 -.022 .098 -.015 32.TECHSUP .016 -.054 -.025 .024 .032 -.025 .018 .020 .042 33.EXPSAVE -.018 -.075 -.014 -.067 -.071 -.051 .011 -.038 -.058 34.EYSLFREL .044 -.041 -.141 .081 .133 -.032 -.129 .078 -.070 35.CNRBLF .059 .034 -.032 .056 .091 -.059 -.053 .030 -.049 36.STATUS -.059 -.027 -.125 .061 .205 .036 -.094 .160 -.084 37.LCLSLFREL -.021 .065 -.119 .087 .221 .033 -.103 .064 -.061 38.ECON .110 -.107 -.062 .038 -.037 -.115 -.096 -.023 .026 39.CNRBHVR .021 .175 -.059 .068 .197 .046 .027 .055 -.030 40.EYCONS .027 .033 .083 -.024 -.023 .007 .083 -.072 .051 Table 38 (cont -) 305 19 20 21 22 23 24 25 26 20.PASBLDMTD .083 21.PRCOST -.017 .026 22.PRLCL -.002 .045 .346 23.PRPEOP -.022 .017 .322 .344 24.PRBLD .035 .080 .226 .246 .178 25.PROPNG -.032 .102 .228 .201 .231 .361 26.GENEVAL -.023 .030 .037 .020 .035 .133 .255 27.RECMEN -.001 .042 .040 .001 .035 .107 .207 .596 28.ACTCOND -.109 .038 .035 .022 .039 .155 .334 .374 29.PASSCOND -.072 .159 -.046 .056 -.005 .086 .015 .198 30.TECEVAL .009 .073 .058 .018 .012 .151 .068 .219 31.ADDTECH .039 .114 -.078 .084 -.039 .088 .054 .213 32.TECHSUP .039 .035 -.017 .004 .004 .119 .184 .160 33.EXPSAVE .028 .056 -.035 .005 -.033 -.O8O .130 -.386 34.EYSLFREL .021 .136 -.059 .064 .003 -.045 -.015 .173 35.CNRBLF -.002 .116 .030 -.004 .022 -.009 .007 .132 36.STATUS .059 .198 -.069 -.034 -.029 -.043 -.055 .095 37.LCLSLFREL .072 .107 -.096 -.092 -.023 -.092 -.100 .129 38.ECON -.010 .015 -.013 .016 .047 .007 .081 .135 39.CNRBHVR .019 .185 -.024 -.063 -.020 -.048 -.070 .104 40.EYCONS .013 .082 .031 .016 -.020 -.011 -.009 -.031 27 28 29 30 31 32 33 34 35 28.ACTCOND .285 29.PASSCOND .163 .192 30.TECEVAL .186 .112 .078 31.ADDTECH .212 .021 .207 .029 32.TECHSUP .117 .155 .061 .056 .068 33.EXPSAVE -.281 .208 .104 -.136 -.121 -.088 34.EYSLFREL .169 .028 .174 -.104 .233 .071 -.030 35.CNRBLF .126 .034 .078 -.024 .106 .044 -.016 .531 36.STATUS .104 .047 .152 .073 .212 .001 -.002 .401 .189 37.LCLSLFREL .131 .002 .236 .018 .290 .061 -.082 .428 .158 38.ECON .158 .085 .051 .131 .046 .070 .010 .441 .198 39.CNRBHVR .091 -.054 .125 .040 .187 -.004 -.068 .214 .306 40.EYCONS -.051 .026 .021 .012 -.045 -.049 .012 -.141 -.146 306 Table 38 (cont.) 36 37 38 39 40 37.LCLSLFREL .272 38.ECON .246 .159 39.CNRBHVR .148 .189 -.093 40.EYCONS -.076 -.075 -.054 -.139 PEARSON CORRELATIONS: NONSOLAR HOMEOWNERS 307 Table 39 Variables l 2 3 4 5 6 8 1.FAMTYPE 2.NUMCHIL .561 3.AGE .288 .364 4.LADDER .021 .033 .001 5.HOUSE .068 .039 .032 .025 6.IMPACT .068 .136 .095 .041 .002 7.AVCOST .148 .182 .173 I.011 .005 .101 8.STATE .038 .053 .013 I.033 .014 .021 .023 9.FEDERAL .122 .077 .128 -.063 .097 .053 .056 .191 10.EDUCATION .093 .071 .122 -.034 .091 .015 .104 .035 .266 11.INCOME .382 .263 .394 I.036 .090 I.012 .223 .019 .294 12.PERSUP .084 .095 .054 I.038 .001 .032 .013 .029 .081 13.GENEVAL .182 .135 .130 I.064 .002 .090 .041 .040 .235 14.EASE .116 .068 .075 I.007 .006 I.006 .014 .041 .132 15.TECEVAL I.001 .025 .033 I.033 .032 -.040 .052 I.015 .143 16.BI .089 .084 .064 -.031 .040 .069 .023 .050 .242 17.INVEST .086 .098 .053 -.068 .008 .068 .017 .060 .018 18.WILLPAY .155 .098 .105 I.011 .024 .020 .024 .014 .144 19.EYSLFREL .102 .068 .109 I.051 .020 .102 .053 .003 .019 20.CNRBLF .075 .086 .097 -.058 .010 .098 .023 .009 .053 21.8TATUS .042 .024 .003 -.019 .005 .056 .025 .077 I.093 22.LCLSFREL .230 .115 .173 I.015 .006 .124 .013 .005 .139 23.ECON .101 .090 .110 -.008 .037 .111 .088 .045 I.062 24.CNRBHVR .097 .124 .079 I.013 .058 .012 .063 .038 .113 25.EYCONS .198 .116 .103 I.057 .002 .123 .073 -.013 .170 Table 38 (cont.) 308 10 11 12 13 14 15 16 17 18 11.INCOME .461 12.PERSUP .111 .127 13.GENEVAL .259 .295 .382 14.EASE .139 .201 .161 .312 15.TECEVAL .104 .058 .144 .309 .167 16.BI .227 .244 .170 .369 .342 .211 17.INVEST .027 .046 .185 .395 .179 .218 .205 18.WILLPAY .179 .250 .201 .382 .199 .165 .230 .233 19.EYSLFREL .008 .122 .122 .284 .122 -.142 .128 .207 .174 20.CNRBLF .101 .149 .172 .329 .121 I.070 .133 .208 .214 21.STATUS .196 .074 .037 .097 .050 -.152 .008 .193 .073 22.LCLSFREL .096 .254 .103 .267 .130 .011 .186 .074 .172 23.ECON .060 .038 .063 .174 .092 .224 .045 .181 .093 24.CNRBHVR .113 .127 .088 .176 .099 .120 .140 .095 .158 25.EYCONS .190 .271 .041 .223 .095 .040 .166 .054 .147 19 20 21 22 23 24 25 20.CNRBLF .636 21.STATUS .466 .280 22.LCLSFREL .359 .252 .140 23.ECON .683 .502 .546 .243 24.CNRBHVR .088 .133 .006 .149 .019 25.EYCONS .133 .182 -.009 .243 .055 .201 309 Table .40 SPP SCALE WEIGHTS: ACTIVE AND PASSIVE TECHNOLOGY OWNERS Variable: Active Passive Energy self reliance .77439 .74226 Local self reliance .58445 .29312 Resource conservation attitudess .74819 .62470 Resource conservation behavior .62811 .67378 Energy conservation behavior .37323 .86689 Personal noneconomic gains .67064 .59034 Personal economic gains .71391 .66532 310 Table 41 COMMUNITY PARTICIPATION: SOLAR AND NONSOLAR HOMEOWNERS SOLAR NONSOLAR Number Number % Answered % Answered Worked on community problems 39 (1477) 23 (464) Contacted politicians 37 (1419) 16 (316) Ran for a political office 3 (96) 1 (18) Wrote letters to editors 18 (689) 7 (146) Worked for a campaign 11 (416) 7 (132) Officer of a community organization 25 (956) 11 (229) Signed a petitian 52 (1964) 39 (789) Gave neighbors advice 49 (1860) 41 (816) None of these 16 (594) 36 (723) 311 Table 42 UNOPERATIVE PERIODS OF SOLAR SYSTEMS System I 1 System I 2 System I 1 System I 2 No. "down” times 25% of samples 1 1 1 1 50% " " 2 2 2 2 75% ” " 3 7 3 5 100% ” ” 400+ 400+ 100+ 100+ X 60 140 6 20 s.d. 222 329 18 38 Median 2 2 2 2 Range 1 I 400+ l I 400+ l I 100+ 1 I 100+ Respondents 1284 122 162 34 Missing data 111 2397 198 847 No. years expected to last 25% of samples 10 10 19 19 50% ” ' 20 20 25 29 75% " ' 20 25 80 100 100% ” " 400+ 300+ 600 600 X 36 42 134 150 s.d. 79 74 219 228 Median 20 20 25 30 Range 1 - 400+ 1 - 300+ 1 - 600+ 1 - 600+ Respondents 2534 354 836 224 Missing data 273 2422 228 835 Operating problems: none few some several many missing Number of 69 20 8 3 1 3% problems 2627) (761 (293) (97) (31) (101) 312 Table 43 THEORETICAL TRIMMED MODELS MANIFEST VARIABLES: SOLAR AND NONSOLAR HOMEOWNERS NONSOLAR SOLAR Variables: 1* 1,399 1,2,3*** 44 4,584 4,5,6: (weights) (weights) ENERGY VULNERABILITY NEEDS Household Type .607 .571 .570 .495 .501 .522 No. of Children I.002 I.004 I.007 I.03O I.037 I.045 Age .543 .564 .565 .493 .503 .509 Future "Energy Crisis“ I.159 I.125 I.125 I.101 I.096 I.005 Dwelling Type .097 .135 .137 .226 .225 .208 Impacts of Energy Crisis .058 I.086 I.086 I.150 I.131 I.039 Energy Costs .170 .234 .236 .475 .462 .429 RESOURCES Problems-Costs Up --.. ---- ---- .075 .072 .011 Problems With Local Policies --.. ---- ---- .055 .051 .001 Problems With People IIII IIII IIII .117 .117 .141 Problems Building Technologies IIII IIII IIII .174 .166 .028 Local Support IIII IIII IIII I.029 I.028 I.015 State Support .034 .027 .029 .088 .089 .117 Federal Support .189 .182 .187 .552 .556 .621 Income .813 .822 .823 .657 .658 .642 Personal Support .377 .364 .347 I.063 I.062 I.019 Solar Group Member IIII IIII IIII .190 I.186 I.039 TECHNOLOGY TYPE Ownership Length IIII IIII IIII .599 .522 I.7o3 Year Active Installed IIII IIII IIII .635 .578 I.690 Year Passive Installed IIII IIII IIII .136 I.072 I.241 Cost of Technology IIII IIII IIII I.023 I.067 I.135 Operating Problems IIII IIII IIII I.592 I.56o .370 EVALUATION General Evaluation .828 .828 .828 .718 .309 .690 Recommended Adoption .572 .572 .572 .671 .288 .655 Condition of Active system ...- ...- ---- .439 .119 .379 Condition of Passive system ---- -... ---- .624 .431 .633 Technical Evaluation .364 .366 .359 .080 I.120 .048 Adding Technologies .648 .649 .647 .581 .350 .626 Technical Support IIII IIII IIII .246 .143 .218 Expected Energy Savings .538 .537 .541 I.397 I.087 I.371 Willingness to Pay More Cost .641 .642 .642 ---- ---- ---- SOFT PATH PREFERENCES Energy Self Reliance .788 .781 .810 .824 .339 .841 Conservation Attitudes .738 .748 .763 .566 .168 .592 Noneconomic Gains .432 .410 .468 .629 .285 .609 Local Self Reliance .642 .637 .607 .701 .381 .681 Economic gains .651 .630 .675 .436 .138 .482 Conservation Behaviors .380 .396 .376 .457 .225 .440 Energy Conservation Behaviors .490 .497 .447 I.228 I.070 I.210 1* I Drop Needs I Evaluation 1,3** I Drop Needs I Evaluation, Resources-SPP 1,2,3*I* I Drop Needs I Evaluation, Resources-SPP, Needs-SPP 48 I Drop Needs I Tech Type 4,544 I Needs Tech Type, Needs-SPP 4,5,6: I Drop Needs I Tech Type, Needs-SPP, Resources-SPP 313 Table 44 THEORETICAL TRIMMED MODELS' - MANIFEST VARIABLES: ACTIVE AND PASSIVE TECHNOLOGY OWNERS Active Passive Variables: 1* 1,28 1,2,3! 1* 1,2# 1,2,3! (weights) (weights) ENERGY VULNERABILITY NEEDS Household Type .490 .484 .485 .063 .139 .064 No. of Children I.074 I.062 -.06l I.205 I.33l I.243 Age .605 .592 .591 .415 .519 .551 Future "Energy Crisis" .135 .154 .154 .087 .107 .146 Dwelling Type .118 .125 .124 .268 .278 .319 Impacts of Energy Crisis .010 I.029 I.025 I.219 I.130 I.033 Energy Costs .319 .340 .338 .787 .714 .695 CONTEXTUAL RESOURCES Problems-Costs Up I.020 I.018 I.019 .121 .117 I.025 Problems With Local Policies I.016 I.011 I.012 .115 .086 I.030 Problems With People .109 .105 .108 I.099 I.085 .041 Problems Building Technologies I.174 I.173 I.176 I.021 I.016 I.063 Local Support I.036 I.039 I.038 I.105 I.lll I.077 State Support .138 .137 .139 I.032 I.03O .033 Federal Support .563 .558 .564 .042 .048 .138 Education .012 .009 .009 -.060 I.062 I.039 Income .645 .653 .646 .933 .941 .912 Solar Group Member I.026 I.027 I.018 I.3O3 I.289 I.33o TECHNOLOGY TYPE Ownership Length .869 .868 .868 .207 .213 .315 Year Active Installed .762 .761 .762 IIII III- III- Year Passive Installed IIII IIII IIII I.006 .001 .110 Cost a: Technology .383 .335 .335 .664 .670 .684 Active Building Method .441 .442 .440 IIII III- ---- Passive Building Method I-II I-I- II-I .880 .876 .843 EVALUATION General Evaluation .741 .742 .744 .728 .726 .732 Recommend Adoption .722 .723 .724 .752 .751 .759 Condition of Active System _ .402 .401 .402 ---- III- III- Condition of Passive System II-I IIII IIII .314 .317 .343 Technical Evaluation I.041 I.038 I.036 .115 .110 .118 Adding Technologies .686 .687 .686 .594 .594 .582 Technical Support .357 .356 .355 .179 .174 .179 Expected Energy Savings I.451 I.452 I.453 I.058 I.052 I.074 SOFT PATH PREFERENCES Ene Self Reliance .856 .852 .851 .829 .823 .821 Conggivation Attitudes .636 .650 .655 .576 .556 .570 Noneconomic Gains .576 .573 .566 .507 .524 .558 Local Self Reliance .615 .608 .605 .649 .669 .610 Economic Gains .584 .567 .568 .562 .563 .619 Conservation Behavior .340 .362 .369 .453 .432 .392 er Conservation En BgKavior I.195 I.218 I.212 I.324 I.300 I.311 1* I Drop Needs I Tech Type 1,2# I Drop Needs I Tech Type, SPP 1,2,3! I Drop Needs I Tech Type, SPP, Resources-SPP 314 Table 45 THEORETICAL TRIMMED MODELS - GOODNESS OF FIT: SOLAR AND NONSOLAR HOMEOWNERS NONSOLAR SOLAR Trimmed relationships 1* 1,3** 1,2,3 44 4,544 4,5,6! BETAS LATENT VARIABLES: Needs I Resources .389 .387 .396 .369 .371 .401 Needs I Evaluation IIII I-II IIII IIII IIII IIII Needs I Tech Type IIII IIII IIII IIII IIII IIII Needs I Spp .165 .175 I-II .017 IIII IIII Resources I Evaluation .473 .472 .470 I.422 I.422 .412 Resources I Spp .042 IIII III- I.175 I.166 IIII Tech Type I Evaluation IIII I-II II-I I.159 I.156 .115 Evaluation I Spp .342 .357 .393 .352 .355 .371 RISQUARE Resources .152 .150 .157 .136 .138 .161 Evaluation .224 .223 .221 .025 .024 .013 Tech Type IIII IIII IIII .178 .178 .170 Spp .188 .184 .157 .156 .156 .138 No. parameters 30 29 28 44 43 42 No. iterations 7 7 6 93 9O 64 Chi-square (no model) 5.141 5.143 5.141 6.725 6.725 6.725 (d.f.) 300 300 300 703 703 703 Chi-square (with model) 2.212 2.178 2.151 4.992 4.981 4.912 (d.£.) 269 269 269 655 655 655 Bentler-Bonnett .570 .576 .582 .258 .259 .270 1* I Drop Needs I Evaluation 1,3** I Drop Needs I Evaluation, Resources-SPP l,2,3*** I Drop Needs I Evaluation, Resources-SPP, Needs-SPP 44 I Drop Needs I Tech Type 4,544 I Drop Needs I Tech Type, Needs-SPP 4,5,6! I Drop Needs I Tech Type, Needs-SPP, Resources-SPP 315 Table 46 TRIMMED MODELS I GOODNESS OP FIT: ACTIVE AND PASSIVE TECRNOLOGY OWNERS Active Passive Trimmed relationships 1* 1,24 1,2,3! 1* 1,24 1,2,3! BETAS LATENT VARIABLES: Needs I Resources .437 .438 .438 .429 .436 .446 Needs I Tech Type IIII IIII IIII IIII IIII IIII Needs I Spp(2) .064 IIII IIII I.127 II-I III- Resources I Tech Type I.4O4 I.403 I.403 I 380 I.381 I.386 Resources I Spp I.058 I.031 IIII I.159 I.216 IIII Tech Type I Evaluation .040 .041 .042 .212 .213 .221 Evaluation I Spp .320 .321 .320 .331 .331 .375 RISQUARE Resources .191 .192 .192 .184 .190 .199 Tech Type .163 .162 .163 .144 .145 .149 Evaluation .002 .002 .002 .045 .045 .045 Spp .107 .103 .102 .197 .188 .141 No. parameters 41 40 39 41 40 39 No. iterations 12 13 12 12 13 13 Chi-square (no model) 6.142 6.142 6.142 7.198 7.198 7.198 (d.f) 435 435 435 595 595 595 Chi-square(with model) 3.551 3.557 3.546 6.033 6.026 5.965 (d.f.) 395 395 395 550 550 550 Bentler-Bonnett .422 .421 .423 .162 .163 .171 1* I Drop Needs I Toch Type 1,24 I Drop Needs I Tech Type and Needs-SPP 1,2,3! I Drop Needs I Tech Type, Needs-SPP, Resources-SPP 316 Table 47 MODIFIED TRIMMED MODELS I MANIFEST VARIABLES: SOLAR HOMEOWNERS SOLAR Variables: 1* 1,24 1,2,3! (weights) ENERGY VULNERABILITY NEEDS Household Type .493 .493 .498 No. 0: Children III- II-I ---- Age .509 .508 .504 Future ”Energy Crisis" IIII IIII ---- Dwelling Type .187 .187 .188 Impacts of "Energy Crisis“ IIII ---- ---- Energy Costs .431 .433 .431 CONTEXTUAL RESOURCES State Support .092 .092 .111 Federal Support .536 .536 .569 Income .712 .712 .675 Personal Support ---- ---- ---- Solar Group Member ---- ---- ---- TECHNOLOGY TYPE Ownership Length .680 .680 .689 Year Active Installed .660 .660 .666 Year Passive Installed .259 .259 .263 Cost or Technologies .321 .321 .299 Active Building Method .474 .474 .466 Passive Building Method I.077 I.077 I.093 PROBLEMS Problems I Costs Up .332 .333 .336 Problems With Local Policies .260 .261 .264 Problems With People .331 .331 .334 Problems Building Technologies .687 .688 .687 Operating Problems .919 .918 .919 EVALUATION General Evaluation .832 .832 .833 Recommend Adoption .759 .759 .759 Condition of Active System .620 .619 .625 Condition of Passive System .337 .335 .329 Technical Evaluation .268 .272 .271 Adding Technologies .355 .354 .343 Technical Support .378 .379 .381 Expected Energy Savings I.507 I.508 I.507 SOFT PATH PREFERENCES Energy Self Reliance .850 .847 .856 Conservation Attitudes .583 .599 .632 Noneconomic Gains .597 .593 .549 Local Selt Reliance .679 .670 .649 Economic Gains .535 .511 .549 Conservation Behavior .384 .411 .407 Energy Conservation Behavior I.233 I.263 I.229 1* I Drop Needs I SPP 1.24 I Drop Needs I SPP and Needs I Tech Type 1,2,3! I Drop Needs I SPP, Needs I Tech Type, and Resources-SPP 317 Table 48 MODIFIED TRIMMED MODELS - MANIFEST VARIABLES: ACTIVE AND PASSIVE TECHNOLOGY OWNERS Active Passive Variables: 1* 1,24 1,2,3! 1* 1,24 (weights) ENERGY VULNERABILITY NEEDS Household Type .495 .495 .453 IIII IIII No. of Children IIII IIII IIII I.226 I.226 Age .594 .579 .606 .579 .574 Future ”Energy Crisis" IIII IIII IIII .126 .144 Dwelling .103 .110 .135 .302 .307 Impacts of ”Energy Crisis" IIII IIII IIII I.112 I.109 Energy Costs .311 .341 .353 .697 .697 CONTEXTUAL RESOURCES State Support .141 .143 .139 IIII I-II Federal Support .585 .583 .583 .018 .010 Income .658 .659 .661 .907 .914 Personal Support IIII IIII III- I.288 I.280 Solar Group Member IIII IIII IIII I.306 I.297 TECHNOLOGY TYPE Ownership Length .882 .882 .878 .319 .226 Year Active Installed .776 .776 .773 IIII III- Year Passive Installed IIII IIII IIII .159 .040 Cost of Technologies .378 .378 .367 .701 .697 Active Building Method .397 .398 .422 IIII I-II Passive Building Method IIII IIII IIII .830 .859 PROBLEMS Problems I Costs Up .385 .384 .384 .276 .277 Problems With Local Policies .333 .333 .332 .530 .530 Problems With People .387 .387 .387 .549 .549 Problems Building Technologies .632 .632 .630 .764 .763 Operating Problems .915 .915 .916 .804 .806 EVALUATION General Evaluation .835 .835 .835 .826 .826 Recommend Adoption .768 .768 .767 .765 .765 Condition of Active System .712 .711 .711 IIII IIII Condition of Passive System IIII IIII IIII .573 .573 Technical Evaluation .231 .233 .232 .252 .252 Adding Technologies .289 .291 .291 .279 .279 Technical Support .381 .381 .381 .285 .285 Expected Energy Savings I.554 I.554 I.554 I.369 I.369 SOFT PATH PREFERENCES Energy Self Reliance .857 .854 .854 .828 .829 Conservation Attitudes .641 .686 .686 .553 .554 Noneconomic Gains .480 .484 .484 .533 .531 Local Self Reliance .574 .568 .568 .658 .658 Economic Gains .705 .646 .646 .573 .573 Conservation Behavior .245 .310 .310 .429 .429 Energy Conservation Behavior I.136 I.232 I.232 I.283 I.282 1* I Drop Needs I SPP 1,24 I Drop Needs I SPP and Needs I Tech Type 1,2,3! I Drop Needs I SPP, Needs I Tech Type, and Resources-SPP 318 Table 49 MODIFIED TRIMMED MODELS I GOODNESS OF PIT: ACTIVE AND PASSIVE TECHNOLOGY OWNERS Active Passive Variables: 1* 1,24 1,2,3! 1* 1,24 1,2,3! BETAS Needs - Resources .426 .427 .427 .439 .440 .431 Needs I Tech Type I.102 I.101 IIII .178 IIII IIII Resources I Tech Type I.348 I.348 I.392 I.477 I.405 I.405 Resources I Spp IIII IIII IIII I.236 I.238 I.262 Tech Type Problems .042 .042 .037 I.003 I.007 I-II Problems I Evaluation .372 .372 .373 .256 .256 .266 Evaluation I Spp .188 .187 .187 .219 .219 .139 RISQUARE Resources .181 .182 .182 .193 .194 .186 Tech Type .162 .162 .154 .185 .164 .164 Problems .002 .002 .001 .000 .000 .000 Evaluation .139 .139 .139 .066 .066 .125 Spp .042 .035 .035 .124 .124 .101 No. parameters 37 36 35 4O 39 39 No. iterations 11 9 9 16 17 22 Chi-square (no model) 5.096 5.096 5.096 6.494 6.494 6.494 (d.f) 435 435 435 528 528 528 Chi-square (with model) 2.686 2.655 2.650 4.962 4.955 4.957 (d.f.) 390 390 390 480 480 480 Bentler'BOnnfltt e‘73 e479 .‘80 e236 e237 e237 1* I Drop Needs :gTech Type 1,24 I Drop Needs I Tech Type and Needs-SPP 1,2,3! I Drop Needs I Tech Type, Needs-SPP, Resources-SPP 319 Table 50 MODIFIED TRIMMED MODELS I GOODNESS OF FIT: SOLAR HOMEOWNERS Solar Variables: 1* 1,24 1,2,3! BETAS Needs I Resources .440 .441 .445 Needs I Tech Type I.177 IIII III- Needs I Spp IIII IIII IIII Resources I Tech Type .475 .403 .407 Resources I Spp I.242 I.243 IIII Tech Type Problems .003 .008 .011 Problems I Evaluation .257 .257 .261 Evaluation I Spp .215 .216 .272 RISQUARE Resources .194 .195 .198 Tech Type .183 .163 .165 Problems .000 .000 .000 Evaluation .066 .066 .068 Spp .126 .126 .074 No. parameters 44 43 42 No. iterations 15 15 15 Chi-square (no model) 7.737 7.737 7.737 (d.f) 666 666 666 Chi-square (with model) 6.085 6.075 6.082 (d.f.) 614 614 614 Bentler-Bennett .214 .215 .214 1* I Drop Needs I SPP 1,24 I Drop Needs I SPP and Needs I Tech Type 1,2,3! I Drop Needs I SPP, Needs I Tech Type, and Resources I SPP BIBLIOGRAPHY BIBLIOGRAPHY Adams, Richard Newbold. Energy and Structure: A Theory 9; Social Power. Austin, Texas: University of Texas Press. 1975. Alford, Robert R. and Roger Friedland. "Political Participation and Public Policy.” Annual Review of Sociology I:429-478. 1975. Anderson, Walter Truett. "Editor's Introduction: The Crisis of Liberalism" in Walter Truett Anderson, ed. Rethinking Liberalism. New York: Avon Books. 1983. Apel, H. and H. Wold. "Soft Modeling with Latent Variables in Two or More Dimensions: PLS Estimation, and Testing for Predictive Relevance." In K. G. Joreskog and H. Wold eds. Systems Under Indirect Observation: Causality, StructureL Prediction Part II. New York: North-Holland Publishing Co. 1982. Areskoug, B. "The First Canonical Correlation: Theoretical PLS Analysis and Simulation Experiments." In K.G. Joreskog and H. Wold, eds. Systems Under Indirect Observation: Causality, StructureL Prediction Part II. New York: North-Holland Publishing Company. 1982. Axelrod, Regina 8. ”Energy Policy: Changing the Rules of the Game.” In Norman J. Vig and Michael E. Kraft, eds. Environmental Policy in the 1980's: Reagan's New Agenda Washington D.C.: Congressional Quarterly Inc. 1984. Barbour, Ian, Harvey Brooks, Stanford Lakoff and John Opie. Energy and American Values. New York: Praeger Publishers. 1982. Bentler, P. M. and Douglas G. Bonett. "Significance Tests and Goodness of Fit in the Analysis of Covariance Structures." Psychological Bulletin 88, 3:588-606. 1980. 320 321 Bentler, Peter M. and George Speckart. "Attitudes 'Cause' Behaviors: A Structural Equation Analysis." Journal 2; Personality and Social Psychology 40, 2:226-238. Bentler, P. M. and David G. Weeks. "Linear Structural Equations with Latent Variables." Psychometrika 45, Bezdek, Roger H., Robert Wending, Gerald E. Bennington, and H. Richard Chew. ”National Goals for Solar Energy: Economic and Social Implications.” Natural Resources Journal 22,2:337-360. 1982. ‘ Bielby, William T. and Robert M. Hauser. "Structural Equation Models." Annual Review g; Sociology 3:137-161. 1977. Bookchin, Murray. "Energy, Ecotechnocracy, and Ecology." In Gary Coates, ed., Resettling America: Ener Ecology and Community. Andover, Massachusettes: Br ck House Publishing Company. 1981. Brody, Charles J. ”Differences by Sex in Support for Nuclear Power." Social Forces 63,1:209-228. 1984. Brooks, Harvey. "Technology: Hope or Catastrophe?" Technology $2 Society 1:3-17. 1979. Brooks, Harvey. "Can Technology Assure Unending Material Progress?" In G. A. Almond, M. Chodorow, and R. H. Pearce, eds.,Prggress and Its Discontents. Berkeley, California: UniVersity of California Press. 1982. Buss, David M. and Kenneth H. Craik. "Contemporary Worldviews: Personal and Policy Implications." Journal g; Applied Social Psychology 13,3:259I280. 1983. Butti, Ken and John Perlin. A Golden Thread:2500 Years 9; Solar Architecture and Technology. Palo Alto, California: Cheshire Books. 1980. Carlson, Richard C., Willis W. Harman, Peter Schwarts, and Associates. Energy Futures, Human ValuesL and Lifestyles: A HE! Look at the Energy Crisis Boulder, Colorado: WestView Press. 1982. Carniero, Robert, ed. Herbert Spencer: The Evolution 9f Society Chicago: University of Chicago Press. 1967. 322 Carpenter, Edwin H. and S. Theodore Chester Jr. "Metropolitan and Nonmetropolitan Homeowner's Awareness and Utilization of Federal Energy Tax Credits: A Western States Perspective.” Paper presented at the Annual Meetings of the Rural Sociological Society, San Francisco. 1982. Coates, Gary, ed. Resettling America: Energy, Ecology and Community; Andover, Massachusetts: Brick House Publishing Company. 1981. Combs, E. Raedene. Part 1: Attitudes of Nebraska Homebuilders Concerni_g the Construction of Solar and Earth Sheltered Homes. Part 2: Attitudes of Nebraska Loan Officers Concerning the Financing of_ Loans for Solar and Earth Sheltered Homes. Lincoln, Nebraska: The Department of Education and Family Resources and the Nebraska Agricultural Extension Service. 1983. Conn, W. David, ed. Energy and Material Resources: Attitudes, Values and Public Policy. AAAS Selected Symposium 75. Boulder, Colorado: Westview Press, Inc. 1983. Cooper, Joel and Russell H. Fazio. ”A New Look at Dissonance Theory.” In Leornard Berkowitz, ed. Advances in Experimental Social Psychology. Volume 17. New York: Academic Press, Inc. 1984. Cooper, Mark N., Theodored L. Sullivan, Susan Punnett, and Ellen Berman. Equity and Energy: Rising Energy Prices and the Living Standards g; Lower Income Americans Boulder, Colorado: Westview Press. 1983. Cose, Ellis. Decentralized Energy Decisions: The Rebirth of Community Power. Boulder, Colorado: Westview Press. 1983. Cottrell, Fred. Energy and Society: The Relation Between Energyy Social Change, and Economic Development. New York: McGraw-Hill Book Company, Inc. 1955. Craig, Paul P. and Mark D. Levine, eds. Decentralized Energy. AAAS Selected Sumposium 72. Boulder, Colorado: Westv ew Press, Inc. 1982. Cronbach. Lee J. "Coefficient Alpha and the Internal Structure of Tests.” Psychometrika 16,8:297I334. 1951. Darley, John M. and James R. Beniger. ”Diffusion of Energy- Conserving Innovations." Journal g; Social Issues 37, 2: 150-171. 1981. 323 Darmstadter, Joel, Joy Dunkerley, and Jack Alterman. "International Variations in Energy Use: Findings From a Comparative Study.” Annual Review 2; Energy 3:201-224. 1978. Davis, David Howard. Energy Politics 3rd ed. New York, New York: St. Martin's Press. 1982. Dietz, Thomas and James P. Hawley. ”The Impact of Market Structure and Economic Concentration on the Diffusion of Alternative Technologies: The Photovoltaic Case." In Max Neiman, ed. 222 Social Constraints 2; Energy Poligy. Lexington. 1982. Dietz, Thomas and Edward L. Vine. ”Energy Impacts of a Municipal Conservation Policy.” Energy 7, 9: 755-758. 1982. Dijkstra, Theo. "Some Comments on Maximum Likelihood and Partial Least Squares Methods." Journal 9; Econometrics 22:67-90. 1983. Duncan, Otis Dudley. Introduction pg Structural Equation Models. New York: Academic Press. 1975. Dunlap, Riley and Marvin E. Olsen. "Hard-Path Versus Soft- Path Advocates: A Study of Energy Activists.” Policy Studies Journal 13: 413-428. 1984. Eastman, Clyde. "Diffusion of Solar Technology in New Mexico". Paper presented at the Annual Meeting of the Rural Sociological Society, San Francisco, California. September 1-4. 1982. Edison Electric Institute. "Soft vs. Hard Energy Paths: A Series of Critical Essays on Amory Lovins 'Energy Strategy: The Road Not Taken?" Electric Perspectives No. 77/3. 1977. "An Introductory Primer for Latent Variable Path Analysis with Partial Least-Squares Estimation." Berkeley, California: University of California Institute of Human Development. Unpublished manuscript. 1986. Falk, R. Frank and Susan Calhoun Stuber. "Introducing Soft Modeling with Partial Least Squares." Paper presented at the 79th Annual Meeting of the American Sociological Association, San Antonio, Texas, August 27-31. 1984. Farhar-Pilgrim, Barbara and Charles T. Unseld. America's Solar Potential: A National Consumer Study. New York, New York: Praeger Press. 1982. 324 Farhar-Pilgrim, Barbara C., Patricia Weis, Charles T. Unseld, and Barbara A. Burns. Public Opinion About Energy: A Literature Review Golden, Colorado:Solar Energy Research Institute U.S. Department of Energy. SERI/TRI53I155. 1979. Farhar, Barbara C., Charles T. Unseld, Rebecca Vories, and Robin Crews. ”Public Opinion About Energy." Annual Review g; Energy 5:141-172. 1980. Fishbein, Martin and Icek Ajzen. Belief, Attitude, Intention and Behavior: AA Introduction pg Theory and Research. Reading, Massachusetts: Addison-Wesley. 1975. Fowles, Jim. "On Chronocentrism." Futures 6,1:65-68. 1974. Garrett-Price, B.A., C.H. Bloomster, and L.L. Fassbender. ”The Economics of Solar, Geothermal and Conventional Space Heating." Alternative Energy Sources 8:43-50. 1980. Gaston, Jerry. "Sociology of Science and Technology." In Paul T. Durbin, ed. A Guide Ag the Culture 2; Science, Technology, 22g Medicine. Chapter 7. New York: Free Press. 1980. Georgescu-Roegen, Nicholas. The Entropy Law and the Economic Process. Cambridge, Massachusettes: Harvard University Press. 1971. Georgescu-Roegen, Nicholas. "Energy and Economic Myths." The Southern Economic Journal 41,3:347-381. Gerlach, Luther. "The Pragmatic Use of the Energy Path Polemic" Paper presented at the workshop of The Social Impacts of Energy Decentralization, the Institute of Man and Science, Department of Energy, June 13-16, 1979. 1981. Gerlach, Luther P. "Dueling the Devil in the Energy Wars." In Paul P. Craig and Mark D. Levine, eds. Decentralized Energy. Boulder, Colorado: Westview Press. 1982. Gilmer, Robert W. The Social Control g; Energy: A Case $9; the Promise g; Decentralized Solar Technologies. Oak Ridge, Tennessee: Oak Ridge Associated Universities. Institute for Energy Analysis. 1980. Goldstein, David B. "Refrigerator Reform: Guidelines for Energy Gluttons." Technolggy Review 86, 2:365-46. 1983. Gouldner, Alvin W. and Richard A. Peterson. Notes 92 Technology and the Moral Order. Indianapolis, Indiana: The Bobb-Merfill Company Inc. Pub. 1962. 325 Greenberger, Martin and Richard Richels. "Assessing Energy Policy Models: Current State and Future Directions." Annual Review g; Energy 4:467-500. 1979. Green, Judy, Steven Emerson, and John Shepard. Lights Out at DOE: How Reagan Has Put America in the Dark About Ener Open Government Project. Washington D. C.: Public Citizen. 1984. Habermas, Jurgen. Toward a Rational Society: Student Egggest, Science and Politics. Boston: Beacon Press. Hamm, Rita R., Steve H. Murdock, and Clarence Parks. "Residential Effects on Energy Use, Conservation and Innovation." Paper presented at the Annual Meetings of the Rural Sociological Society, San Francisco, August. 1982. Hirst, Eric, R.W. Barnes, S.M. Cohn, K. R. Corum, D. L. Greene, D. M. Hamblin, G. Kulp, D. B. Reister, and T.A. Vineyard. Energy Use from 1973 to 1980: the Role of Improved Energy Efficiengy. Oak Ridge, Tennessee: Oak Ridge National Laboratory. Institute for Energy Analysis. ORNL/CON-79. 1981. Hirst, Eric, Robert Marlay, David Greene, and Richard Barnes. "Recent Changes in U.S. Energy Consumption: What Happened and Why." Annual Review g; Energy 8:193- 245. 1983. Holdren, John P. ”Decentralized Renewables: The Environmental Dimension." In Paul P. Craig and Mark D. Levine, eds. Decentralized Energy. AAAS Selected Symposium 72. Boulder, Colorado: Westview Press Inc. 1982. Holdren, John P., Gregory Morris, and Irving Mintzer. ”Environmental Aspects of Renewable Energy Sources." Annual Review 25 Energy 5:241-291. 1980. Hornback, Kenneth. ”Orbits of Opinion: the Role of Age in the Environmental Movement's Attentive Public, 1968- 1972." Ph.D. dissertation. Michigan State University. 1974. Hornick, William F. and Gordon A. Enk. Value Issues in Technology Assessment: How to Understand and Deal _with Them in the Decentralized Solar Energy Technology Assessment Program. Washington D. C.: Division of Planning and Technology Transfer. Department of Energy. EX-77-C-Ol-4120. 1980. 326 Inhaber, H. "Risk with Energy from Conventional and Nonconventional Sources." Science 203:718-723. 1979. James, Lawrence R., Stanley A. Mulaik, and Jeanne M. Brett. Causal Analysis: Assumptions, Models and Data Beverly Hills, California: Sage Publications. 1982. Joreskog, Karl G. "Structural Equation Models in the Social Sciences:Specification, Estimation and Testing.” In Karl G. Joreskog and Dag Sorbom, eds. Advances 13 Factor Analysis and Structural Equation Models. Cambridge, Massachusetts: Abt Books. 1979. Joreskog, K. G. and Herman Wold. ”The ML and PLS Techniques for Modeling with Latent Variables." In K. G. Joreskog and Herman Wold, eds. Systems Under Indirect Observation: Causalityy StructureL Prediction. Part I. New York: North-Holland Publishing Co. 1982. Just, James and Lester Lave. "Review of Scenarios of Future U.S. Energy Use.” Annual Review 93 Energy 4:501-536. 1979. Kahn, Edward. "The Compatibility of Wind and Solar Technology with Conventional Energy Systems." Annual Review 2; Energy 4:313-352. 1979. Kapstein, Ethan Barnaby. ”A History of Solar Energy Research in the United States." M.A. thesis. University of Toronto. 1978. Katz, Alfred M. ”Self-Help and Mutual Aid: An Emerging Social Movement?" Annual Review 23 Sociolggy 7:129-155. 1981. Keating, Kenneth M. "The Social and Attitudinal Determinants of the Behavioral Intention to Adopt Solar Energy Systems." Ph.D. dissertation. Washington State University. 1982. Keating, Kenneth M. and Eugene Rosa. ”Policy Relevant Variables in Homeowners' Decision to Adopt Solar Energy Systems." Paper presented at the American Council for an Energy-Efficient Economy conference in Santa Cruz, California, August. 1982. Keating, Kenneth M., Eugene Rosa, and Barbara C. Farhar- Pilgrim. "Predicting Intentions to Adopt Solar Energy Technologies with the Fishbein Model: Tests of Robustness and Sufficiency". Pullman, Washington: Department of Sociology, Washington State University. Unpublished paper. 1982. 327 Kempton, Willett and Laura Montgomery. "Folk Quantification of Energy." Energy 2 The Internation Journal (January) 1983. Kempton, Willett, Craig K. Harris, Joanne G. Keith, and Jeffrey S. Weihl. "Do Consumers Know 'What Works' In Energy Conservation?" Paper presented at the 1982 Summer Study, American Council for an Energy-Efficient Economy. Santa Cruz, California. 22-28 August. 1982. Kenny, David A. Correlation and Causality. New York: John Wiley and Sons. 1979. Kim, Jae-On and Charles M. Mueller. Factor Analysis: Statisticsl Methods and Practical Issues. Beverly Hills, Calif.: Sage Publications. 1978. Kinsley, Michael J. "Economic Renewal Project." In Nebraska Energy Office, ed. Community Energy Management 3g an Economic Development Strategy. Proceedings of a National Colloquium sponsored by the University of Nebraska-Lincoln and by the Nebraska Energy Office, Lincoln, Nebraska, October 14-16. 1984. Kowalski, Robert and Bill McBean. "City Enacts Wood-burning Ban." The Denver Post, Oct. 28, p. l, 1986. Leonard-Barton, Dorothy. "The Diffusion of Active Residential Solar Energy Equipment in California." In A. Shama, ed. The Diffusion g; Solar Energy Innovation. New York: Praeger Press. 1981. Levine, Mark S. Canonical Analysis and Factor Comparison. Beverly Hills, California: Sage Publications Inc. Sage Univeristy Paper Series, on Quantitative Applications in the Social Sciences No. 07-006. Levine, Mark D. and Paul P. Craig. "A Decade of United States Energy Policy." Annual Review g; Energy 10:557- 587. 1985. Lodwick, Dora and Denton E. Morrison. "Appropriate Technology." In Don A. Dillman and Daryl J. Hobbs, eds. Rural Society in the U.S.: Issues for the 1980's. Boulder, Colorado: WestView Press. 1982. Lohmoller, Jan-Bernd. LCPLS 1.6 Program Manual: Latent Variable Path Analysis with Partial Least-squares Estimation. Koln, Germany: Zentralarchiv fur Empirischesozialforschung. Universitat zu Koln. 1984. 328 Lonnroth, Mansl Peter Steen and Thomas Johansson. Energy in Transition: A Report on Energy Policy and Future Options. Berkeley, California: University of California Press. 1980. Lovins, Amory. "Energy Strategy: The Road Not Taken?" Foreigg Affairs 55, 1:65-96. 1976. Lovins, Amory. Soft Energy Paths; Toward g Durable Peace New York: Harper and Row. 1977. Lovins, Amory B. ”Soft Energy Technologies." Annual Review 93 Energy 3:477-517. 1978. Lovins, Amory. "Economically Efficient Energy Futures.” In Bach et al. Energy/Climate Interactions. Dordrecht, Netherlands: Reidel. 1980. Lovins, Amory B. and L. Hunter Lovins. "The Writing on the Wall." In Robert W. Haigh, George Gerbner, and Richard B. Byrne, eds. Communications in the Twenty-First Century New York: John Wiley and Sons. 1981. Lovins, Amory and L. Hunter Lovins. Brittle Power: Energy Strate for National Security. Andover, Massachusetts: Brick House Publishing House. 1982. Lovins, L. Hunter, Ken Anderson and Amory B. Lovins. "Energy and Community." In Walter Truett Anderson, ed., Rethinking Liberalism. New York: Avon Books. 1983. Mannheim, Karl. Freedom Power and Democratic Planning. New York: Oxford University Press. 1950. Marcuse, Herbert. One-Dimensional Man: Studies $2 the Ideolggy g; Advanced Industrial Society. Boston: Beacon Press. 1968. McFarland, Andrew 8. ”Energy Lobbies." Annual Review g; Energy 9:501-527. 1984. McKie, James W. "Federal Energy Regulation." Annual Review 2; Energy 9:321-349. 1984. McStay, Jan R. and Riley E. Dunlap. ”Male-Female Differences in Concern for Environmental Quality." International Journal g; Women's Studies 6,4:291-301. 1983. Merton, Robert K. Science, Technology and Society in Seventeenth Century England. New Jersey: Humanities Press. 1970. 329 Mesthene, Emmanuel G. Technological Change: Its Impact 93 Man and Society. Cambridge, Massachusettes: Harvard University Press. 1970. Milbrath, Lester W. Environmentalists: Vanguard for 3 New Society. Albany, New York: State University of New York Press. 1984. Mills, C. Wright. The Sociol ical Imagination New York: Oxford Univers ty Press. 1959. Mills, C. Wright. Powerl Politics and People: 223 Collected Essays 22 E; Wright Mills. Ed ted by Irving Louis Horowitz. New York: Oxford University Press. 1963. Mitchell, Robert C. Public Opinion on Environmental Issues: Results g; a National Public Opinion Survey. Washington, D.C.: Council on Environmental Quality. 1980. Mitchell, Robert Cameron. "Public Opinion and Environmental Politics in the 1970's and 1980's." In Norman J. Vig and Michael E. Kraft, eds. Environmental Policy in Egg 1980's: Reagan's New Agenda. Washington, D.C.: Congressional Quarterly Inc. 1984. Morris, David. Self-Reliant Cities: Energy and the Transformation 95 Urban America. San Francisco: Sierra Club Books. 1982. Morrison, Denton. "The Soft, Cutting Edge of Environmentalism: Why and How the Appropriate Technology Notion is Changing the Movement." Natural Resources Journal 20,2:275-298. 1980. Morrison, Denton E. ”Soft Tech/ Hard Tech, Hi Tech/Lo Tech: A Social Movement Analysis of Appropriate Technology." In Gene F. Summers, ed. Technology and Social Change i2 Rural Areas. Boulder, Colorado: Westview Press. 1983. Morrison, Denton. "How and Why Environmental Consciousness Has Tricked Down." In Allan Schnaiberg, Nicholas Watts and Klaus Zimmerman, eds. Distributional Conflicts i2 Environmental-Resource Policy. New York: St. Martin's Press. 1986. Morrison, Denton E. and Riley E. Dunlap. ”Is Environmentalism Elitist?” In Frederick H. Buttel and Craig E. Humphreys, eds. Environment Energy and Society. University Park, Pennsylvania:Pennsylvania State University Press. 1985. 330 Morrison, Denton E. and Riley E. Dunlap. "Environmentalism and Elitism: A Conceptual and Empirical Analysis." Environmental Management 10, 5:581-589. 1986. Morrison, Bonnie Maas and Peter Michael Gladhart. "Energy and Families: The Crisis and the Response". Journal 9; Home Economics (January):15-18. 1976. Morrison, Denton and Dora G. Lodwick. ”The Social Impacts of Soft and Hard Energy Systems: The Lovins' Claims as a Social Science Challenge." Annual Review 2; Energy 6:357-378. 1981. Morrison, Denton. Forthcoming. "Appropriate Technology: the Movement, the Paradigm, the Source, the Impact." In Craig K. Humphrey, Frederick H. Buttel, and Annette Frahm, eds. Environment and Society. University Park, Pennsylvania: Pennsylvania State University. Forthcoming. Mumford, Lewis. Technics and Civilization. New York: Harcourt, Brace and Company. 1934. Mumford, Lewis. The Myth 95 the Machine: Technics and Human Development. New York: Harcourt, Brace and Company. 1967. Nash, Hugh, ed. Energy Controversy: Soft Path Questions and Answers By Amory Lovins and His Critics. San Francisco: Friends of the Earth. 1979. Nebraska Energy Office. Community Energy Management 35 an Economic Development Strategy. Proceedings of a National Colloquium sponsored by the University of Nebraska- Lincoln and by the Nebraska Energy Office, Lincoln, Nebraska, October 14-16. 1984. Nelson, Lynn D. and Julie A. Honnold. "Decentralized Energy Technology and Diffusion Research: The Case of Passive Solar Design." Paper presented at the 75th Annual Meeting of the American Sociological Association, New York, New York. August 27-31. 1980. Newman, Dorothy. "Energy Consumption in the Life Cycle." Paper presented at the 77th Annual Meeting of the American Sociological Association, San Francisco, September 6-10. 1982. Odum, Howard T. Environment, Power, and Society. New York: Wiley-Interscience. 1971. Ogburn, William F. Social Change with Respect pg Culture and Original Nature. New York: B. W. Huebsch. 1922. 331 Ogburn, William F. Machines and Tomorrow's World. New York: The Public Affairs Committee. 1938. Olsen, Marvin E. ”Public Acceptance of Consumer Energy Conservation Strategies." Journal 2; Economic Psychology 4:183-196. 1983. Olsen, Marvin E., Eugene A. Rosa, and Don A. Dillman. ”The Public Speaks on Nuclear Energy and Energy Conservation.” Paper presented at the Annual Meeting of the American Association for the Advancement of Science, Washington D.C., January. 1982. Perelman, Lewis J. ”Speculations on the Transition to Sustainable Energy". Ethics 90,3:392-416. 1980. Perelman, Lewis J., August W. Giebelhaus, and Michael D. Yokell. Energy Transitions: Long-Term Perspectives. AAAAS Selected Symposium 48. Boulder, Colorado: Westview Press. 1981. Perrow, Charles. Normal Accidents: Living with High-Risk Technologies. New York: Basic Book Inc. Pub. 1984. Piernot, Craig A., Mary Alice Rothweiter, Alice Levine, and Robin Crews. Likely Social Impacts 93 Proposed National-Level Policy Initiatives. Golden, Colorado: Solar Energy Research Institute. SERI/TR-743-786. 1981. Price, Derek J. de Solla. Little Sciences, Big Sciences. New York: Columbia University Press. 1963. Purdy, Dean A. "The Corporate Energy Elite: An Expanding Basis of Influence." Paper presented at the 80th Annual Meeting of the American Sociological Association, Washington D.C. August 26-30. 1985. Reece, Ray. The Sun Betrayed: A Repon Ln the Corporate Seizure Lf U. 8. Solar Energy Development. Boston, Massachusetts: South End Press. 1979. Reed, John and John M. Wilkes. "Sex and Attitudes Toward Nuclear Power.” Paper presented at the Annual Meetings of the American Sociological Association, August. 1980. Ridgeway, James. Powering Civilization: The Complete Energy Reader. New York: Pantheon Books. 1982. Rifkin, Jeremy. Entropy: A New World View. New York: The Viking Press. 1980. 332 Roessner, J. David. "U.S. Government Solar Policy: Appropriate Roles for Nonfinancial Incentives." Policy Sciences 15:3-21. 1982. Rogers, Everett M. Diffusion 9g Innovations. 3rd edition. New York: Free Press. 1983. Rogers, Everett and Floyd Shoemaker. Communications Lf Innovations 2nd edition. New York: Free Press. 1971. Rosa, Eugene A. and Gary E. Machlis. "Energetic Theories of Society: An Evaluative Review.” Sociological Inquiry 53, 2/3 (Spring):152-178. 1933. Roszak, Theodore. ”America After Affluence: an Era of Limits and Values." In Walter Truett Anderson, ed. Rethinking Liberalism. New York: Avon Books. 1983. Rybczynski, Witold. Paper Heroes: a Review LfA _ppropriate Technologies. Garden City, New York: Anchor Press/Doubleday. 1980. Sackett, James G. "Climate-Appropriate Superinsulation: A Key Strategy for Affordable Housing in the Future." St. Louis, Missouri: Saint Louis Energy Management Program. Office of Mayor. 1984. Schipper, Lee. "Progress in More Efficient Energy Use: The Importance of Keeping Score." Paper presented at the Annual Meeting of the American Association for the Advancement of Science, January 5, Washington, D.C. 1982. Schipper, Lee and J. Lichtenberg. "Efficient Energy Use and Well Being: The Swedish Example." Science 194:1001- 1013. 1976. Schnaiberg, Allan. The Environment: From Surplus Lo Scarcity. New York: Oxford University Press. 1980. Schnaiberg, Allan. "Soft Energy and Hard Labor? Structural Restraints on the Transition to Appropriate Technology." In Gene Summers, ed. Technology and Social Change ip Rural Areas. Boulder, Colorado: Westview Press. 1983. Schnaiberg, Allan. ”Redistributive Goals versus Distributive Politics: Social Equity Limits in Environmental and Appropriate Technology Movements." Sociological Inquiry 53, 2/3:200-219. 1983a. Schumacher, E. F. Small is Beautiful. Economics Ls if People Mattered. New York: Perennial Library. 1973. 333 Schuman, Howard and Michael Johnson. ”Attitudes and Behavior." Annual Review 2; Sociolpgy 2:161-207. 1976. Schurmann, Franz. ”Ideology at the Grass Roots: Community, Capital, Cosmology.” In Walter Truett Anderson, ed. Rethinking Liberalism. New York: Avon Books. 1983. Schurr, S. H. ”Energy Use, Technological Change, and Productive Efficiency: An Economic-Historical Interpretation.” Annual Review 2; Energy 9:409-425. 1984. Shippee, Glenn. ”Energy Consumption and Conservation Psychology: A Review and Conceptual Analysis." Environmental Management 4, 4: 297-314. 1980. Solar Energy Research Institute. A New Prosperity: Building 3 Sustainable Ener Future. Andover, Massachusetts: Brick House Publish ng. 1981. Sorokin, Pitirim. Social and Cultural Dynamics. Vol I-IV New York: American Book Co. 1937-1941. Stern, Paul C. and Elliot Aronson, eds. Energy Use:The Human Dimension. New York: Freeman and Co. 1984 . Stout-Wiegand, Nancy and Roger B. Trent. "Sex Differences in Attitudes Toward New Energy Resource Development." Rural Sociology 48,4:637-646. 1983. Swisher, Joel N. "Measured Performance of Passive Solar Buildings." Annual Review g; Energy 10:201-216. 1985. Thurow, Lester C. The Zero-Sum Society; Distribution and the Possibilities for Economic Change. New York: Basic Books, Inc. Publishers. 1980. Totten, Michael. The Road pp Trillion Dollar Energy Savings: 5 Safe Energy Platform. Public Citizen's Critical Mass Energy Project. Washington D.C.: Public Citizen. 1984. Unseld, Charles and Robin Crews. Residential Solar Energy Users: 5 Review pf Empirical Research and Related Literature. Golden, Colorado: Solar Energy Research Institute (SERI/ TR 354-245). 1980. U.S. Department of Commerce. Bureau of the Census. "United States Summary." 1980 Census 9; Population: General Social and Economic Characteristics. PC80-l-C1. Washington, D.C.: Government Printing Office. 1983. 334 Van Liere, Kent and Riley E. Dunlap. ”The Social Bases of Environmental Concern: A Review of Hypotheses, Explanations and Empirical Evidence." Public Opinion Quarterly 44,2: 181-197. 1980. Veblen, Thorstein. Im mperial Germany and the Industrial Revolution. New York: The Viking Press. 1939. Vine, Edward. "Planning for An Energy Conserving Society: The Davis Experience." Report LBID-165, February. Lawrence Berkeley Laboratory. 1979. Vine, Edward. "Solar Energy: Two Sociopolitical Perspectives." Ph.D. dissertation. University of California, Davis. 1980. Vine, Edward L. ”Solar Energy:a aSocial Movement Analysis of an Alternative Technology". From Solar Ener Two Sociopolitical Perspectives Ph. D. dissertat on. University of California at Davis, Davis. 1980a. Walker, James A. Dr., Theador N Rauh, and Karen Griffin. ”A Review of the Residential Conservation Program." Annual Review 2; Energy 10:285-315. 1985. Warkov, Seymour. Solar Adopters and Near-Ado ters: A Study Lf the HUD Hot Water Grant Program. Cambridge,- Massachusetts: Northeast Solar Energy Center. 1979. Weinstein, Jay. Sociology/Technology: Foundations 2; Postacademic Social Science. New Brunswick: Transaction Books. 1982. White, Leslie A. The Evolution Lf Culture: The Development of Civilization to the Fall of Rome. New York: McGraw- Williams, Robert H., Gautam S. Dutt, and Howard S. Geller. "Future Energy Savings in U.S. Housing." Annual Review pp Energy 8:269-332. 1983. Wold, Herman. "Model Construction and Evaluation: When Theoretical Knowledge is Scarce." In Jan Kmenta and James B. Ramsey, eds. Evaluation Lf Econometric Models New York: Academic Press. 1980. Wold, H. "Soft Modeling: The Basic Design and Some Extensions.” In K. G. Joreskog and H. Wold, eds. Systems Under Indirect Observation: Causality, Structure, Prediction Part II. New York: North-Holland Publishing Co. 1982. 335 Yankelovich, Daniel and Bernard Lefkowitz. "The Public Debate on Growth: Preparing for Resolution." Technological Forecasting and Social Change 17:95-140. 1980.