Choice determinants for the (non) adoption of energy efficiency technologies in households
A literature review
Elvira Moukhametshina
Environmental Management in the Swedish Manufacturing Industries
© You may use the contents of the IIIEE publications for informational purposes only. You may not copy, lend, hire, transmit or redistribute these materials for commercial purposes or for compensation of any kind without written permission from IIIEE. When using IIIEE material you must include the following copyright notice: ‘Copyright © Elvira Moukhametshina, Luis Mundaca, Lena Neij, IIIEE, LundUniversity. All rights reserved’ in any copy that you make in a clearly visible position. You may not modify the materials without the permission of IIIEE.
Published in 2008 by IIIEE, LundUniversity, P.O. Box 196, S-221 00 LUND, Sweden,
Tel: +46 – 46 222 02 00, Fax: +46 – 46 222 02 10, e-mail:
Printed by KFS AB, Lund.
ISSN 1650-1675
Acknowledgements
This report was developed through the vast body of literature that had been evolving over the last four decades. I would like to extend my gratitude to Mses. Mithra Moezzi, Lisa Skumatz, Diana Uitdenbogerd and Françoise Bartiaux for being open and kindly advising me in the very beginning of this review. Not the least I appreciate their practical help in providing me with some key references and publications that helped to further structure this report.
Messrs. Harold Wilhite and Richard Wilk have also been very helpful in providing access to some of their key publications.
Particular thanks belong to Professor Lena Neij whose structured and clear vision as well as academic and professional experience made, first of all, the project possible in the time when it is of the essence. Her supervision allowed posing the right questions and looking for objective answers, and guided this report.
Finally, I would like to thank Luis Mundaca for contributing with revision and streamlining the finalisation.
Lund, October 2008
Elvira Moukhametshina.
Table of Contents
List of Figures
List of Tables
List of Appendices
1.Introduction
2.Analytical frameworks to approach the (non) adoption of energy efficiency technologies
2.1Co- (or non-energy) benefits approach
2.2Intervention factors approach
2.3Innovation-related studies
2.4Implicit discount rates
2.5Others approaches
3.Determinants induced by energy efficient technologies
3.1Price
3.2Operating costs
3.3Time
3.4Status/visibility/appearance
3.5Comfort
3.6Branding / design
3.7Compatibility
3.8Performance
3.9Observability
3.10Complexity
3.11Choice/problem solving
4.Demographical factors
4.1Age
4.2Household size and/or composition
4.3Gender
5.Factors related to buyer’s characteristics
5.1Income
5.2Knowledge/information/awareness
5.3Attitude/environmental consciousness
5.4Education/occupation
5.5Type of dwelling
5.6Life style
5.7Timing / Appropriate moment
6.Contextual factors
6.1Ownership/split incentive problem
6.2Intermediaries
6.3Policy incentives
7.Energy (efficiency) technologies
7.1Space conditioning and building envelope
7.2Sanitary hot water
7.3Renewable energy technologies
7.4Lighting
7.5Consumer appliances
8.Concluding remarks
List of Figures
Figure 1. Comparison of U.S. and Thai consumer priorities.
Figure 2. Main factors preventing increased saturation of CFLs in the households with CFLs. Northwest Pacific. 2006.
Figure 3. Main factors preventing increased saturation in the households without CFLs. Northwest Pacific. 2006.
Figure 4. Distribution of the external influence in the adoption of consumer durable products.
List of Tables
Table 1: Main drivers for adopting energy related measures. Results of the on-line survey.
Table 2: Estimated NEBs within EnergyStar New Homes programme.
Table 3: Estimated NEBs within EnergyStar Homes Performance programme.
Table 4: Reasons to buy and not to buy energy-efficient appliances (in %)
Table 5: Rank order of insulation by men and women
Table 6: Demographic and attribute perception ratings
Table 7: Effect of market factors on CFL sales in California in 2005.
Table 8: Ownership of CFLs according to social-demographic characteristics.
Table 9: Main reasons for obtaining a CFL among owners. Cultural survey.
Table 10: Non-energy benefits / impacts of household appliances. Percent of overall non-energy benefits.
Table 11: Overall effectiveness of energy labelling.
Table 12: Differences among adopter categories (vitroceramic hobs).
Table 13: Effect of market factors on resource efficient washing machines sales in California (US).
List of Appendices
Appendix 1:Bibliography reviewed but not referenced ...... 57
Appendix 2:Bibliography not reviewed...... 3
1
1.Introduction
This study presents a literature review focused on energy efficiency and determinants of the (non) adoption of energy efficient technologies in the household sector. The key guiding question is what determinant should be taken into account when analysing future energy (service) demand and potentials of reducing future energy demand by the use of different energy policy instruments. Based on a literature review, the objective of this report is to identify (mostly) quantitative studies that look at the criteria that determine the (non) adoption of energy efficient technologies in the household sector.[1]The report highlights rational economic considerations complemented with co-benefits and behavioural aspects.
Low investment in energy efficient technologies is often identified as a result of the ‘energy efficiency gap’ (see e.g. Jaffe and Stavins, 1994a; Stern and Aronson, 1984; Weber, 1997). This term attempts to capture the slow diffusion of profitable energy efficient technologies that fail to achieve market success. The energy efficiency gap is described in terms of market failures and barriers, indicating that the investors do not choose energy technologies although they are cost effective. In addition, it could be argued that consumer decisions do not respond to the model of rational choice behaviour. Early work done by Lutzenhiser (1992) spotted evidences of lack of economic rationality in consumer decisions to forego some obviously energy efficient measures. In fact, one can safely argue that the approach of economic rationality is inadequate to properly reflect technological consumer preferences. Investment costs are only part of a great variety of variables that frame and drive energy related consumer’s investment decisions. For instance, design, comfort, equipment’s brand, timing, functionality, reliability, learning, marketing, environmental awareness, etc., are likely do influence altogether the decision about an energy-technology choice/purchase. From the societal point of view e.g. environment, the investment outcome is likely to be unsatisfactory so there is a great need in public policy to better understand consumer investment decisions in the context of energy use.
The literature review presented here is an attempt to picture the whole spectrum of determinants and to highlight those playing the strongest roles in households’ adoption decision-making within a range of energy (efficient) technologies. The objective is to capture studies focusing on qualitative as well as quantitative aspects of investments determinants. The determinants have been described form different perspectives:
- Determinants induced by energy efficient technologies
- Demographic aspects that affect (non) adoption
- Factors related to buyer’s characteristics that affect the determinants of investments
- Contextual factors that affect the decision-making process of (non) adoption
The outline of this report is as follows. Chapter 2 is devoted to a short overview of earlier meta- studies and reviews in the field of investments determinants of energy efficiency technologies. This report highlights different analytical approaches to investigate determinants of the (non) adoption of energy efficiency technologies. In Chapter3 the report addresses the identified criteria or determinants of (non) adoption decisions based features of technologies. Cases of determinants are provided in terms of geographical location of samples, observed results and methods by which the results were obtained. In Chapter 4, 5 and 6 the report describes the (non) adoption of efficient technologies in terms of demographic aspects, the characteristics of (potential) adopter/households and the contextual factors in which the (non) adoption process can take place. In Chapter 7, takes a different approach, as it looks into specific technologies and describe the most important determinants on a technology-basis; highlighting that different determinants will be of different importance for different technologies. Finally, in Chapter 8 some concluding remarks are drawn.
2.Analytical frameworks to approach the (non) adoption of energy efficiency technologies
Research in the area of choice determinants for in the adoption of energy efficiency technologies is not novel. Hirst and Goeltz (1985:25) state that “the critical determinants of the household decisions to retrofit are assumed to be the capital and operating costs of the retrofit choices”. This is generally the case of conventional wisdom for various household appliances. However, the literature on several determinants affecting energy (efficiency) technologies in the household sector is vast. In fact, approaches to identify and/or quantify the choice determinants have been developed over the years. Several studies have been published as well as a number of reviews addressing this topic, in particular looking at determinants outside pure financial aspects (see e.g. Lutzenhiser, 1993; Stern, 1986; Wilhite et al., 2000; Uitdenbogerd, 2007).
2.1Co- (or non-energy) benefits approach
To begin with, several research efforts are identified in the literature when it comes to co-benefits (e.g. improved housing comfort level, reduced noise, etc.) of energy efficiency technologies influencing their adoption. For instance, Mills and Rosenfeld (1996) note that many co-benefits (also called ‘non-energy benefits’) play critical role in consumer perception (and adoption) of energy-related technologies. Likewise, yet in the early 80s Stern and Aronson (1984:62) noted that “people do not usually weigh the potential value of the energy saved by one purchase against the pleasure, convenience or status achievable by alternative purchases”.
Stoecklein and Skumatz (2007) have weighted co-benefits for four various technologies associated with residential energy efficiency initiative (New Zealand, Zero and Low Energy Homes). Both positive and negative impacts (benefits/losses) have been reviewed. It is suggested, “residents place considerable value on the lifestyle benefits from energy-efficiency features of their homes, beyond benefits from energy savings”. The authors observed that energy related technologies have a potential to bring benefits other than energy saving (e.g. reduced noise, increased comfort, better energy bill control, etc.).[2] Significant benefits may relate to lifestyle and natural environment.[3] This type of non-energy impacts either becomes a “component of decision-making” or a “contributing reason for satisfaction”. Stoecklein and Skumatz (2007:1962) continue with noting that co-benefits are, indeed, market goods that influence the adoption of energy efficiency technologies as “they are purchased by consumers bundled with the energy-efficiency appliances that produce them”. The authors refer to ‘motivation’ factors that were classified earlier by Lutzenhiser (2006:90), namely: specific system/building concern; environmental health and energy costs; comfort level; and resource conservation. Amann (2006) stresses data collection and work to develop for a proper methodology for incorporating co-benefits in cost-benefit analysis. At earlier stage Skumatz (2002:307-316) compared use of three methods to estimate ‘hard-to-measure’ non-energy benefits of participants of low-income weatherisation programmes. Skumatz (2002) concluded that estimation of overall non-energy benefits in relation to energy benefits within weatherisation programmes could range from 80% to 100% or USD65-USD100.
- Relative valuation. The following benefits have been reported as more valuable than energy savings according to this approach: Control of bills (by 52% of respondents); comfort (34%); environmental (17%); maintenance (16%); moving avoided (13%); change in number of sick days (8%); appearance (6%); added features (6%); noise reduction (4%).
- Willingness-to-pay. This evaluation method approached what programme participants care about. The following results were obtained: comfort (76%); education/control (55%); features/options (30%); noise reduction (30%); appearance (29%). Yet, this resulted in overstatement of individual non-energy benefits that generated a bigger sum than participants would be willing to pay for overall benefits (including energy benefits).
- Labelled Magnitude Scaling. This approach brought about the figure close to the first method. Non-energy benefits comprised 99% of energy benefits from weatherisation and house envelope improvement. In dollar terms value of non-energy benefits was computed to USD70-USD110.
Within the context of co-benefits, Knight et al. (2006) bring forward that motivational surveys suggest that customers incorporate “perceived non-energy benefits” into their decisions, which does not exclude their rational choice. The benefits that accompany retrofitting choices (insulation, appliances, etc.) include increased comfort, reduced noise, improved health, safety, durability properties, and a sense of environmental citizenry, first-on-the-block status, long-term value, and overall peace in mind. They may have a dominating influence on decisions compared to the energy (operational) costs, use and conservation aspects as such. They conclude that based on their survey’s tentative results, “most homeowners appear to value the variety of non-energy benefits much more higher than the energy cost savings”. (Knight et al., 2006: 5-8).
While some authors did not place focus on decision-making phase, for others it seemed central. Knight et al. (2004) do highlight the existence of “full range of performance benefits” that helps homeowners to justify their (investment or purchase) decisions when “energy efficiency alone is insufficient”. (Knight et al., 2004: 7-162). They note that such other benefits would have their subjective utility being characteristic to individual homeowners. Such benefits like “family health, safety, comfort, or prestige” may be significant to become a factor in a purchase or investment decision. Further on the authors share their reservation on available dependable statistical data or studies underway to shed some light on measurable co-benefits effects or perceptions (though rather on the retrofitting part).[4]
2.2Intervention factors approach
Another important systematic approach looking into energy efficiency intervention success factors was taken by Uitdenbogerd et al. (2007).[5] Their review covered vast body of research in various energy efficiency measures beyond housing envelope and weatherisation. It aimed to identify the determinants influencing different components of end-use energy efficiency and present their significance. The approach was to estimate importance of determinants, including the investment behaviour of households with the help of Intervention Mapping Protocol (normally applied in healthcare). This evaluation is basically derived from the number of times respondents in the reviewed references were revealing statistical significance of one or another determinant in a specific context. The team has arrived at the estimations of importance of various criteria on household investment behaviour: They identified personal and external determinants, demographical and contextual factors and quantified their influence/strength with the help of Intervention Mapping Protocol.
Based on the results of their intervention success quantification attempt, the strongest degree of influence is attributed to knowledge, information / learning need / cost estimation, product use (see also Stern et al., 1986a; Spapen, 2003; Derijke and Uitzinger, 2004; Derijke et al., 2001). Uitdenbogerd et al. (2007) regarded a weaker degreeof influence as factors that play a role. Here, a number of studies point several factors that play (a weak) role in investment behaviour, namely: appearance, status, visibility, symbolic value, leisure; optimising luxury, price-judgement/quality, problem solving, service, appropriate moment (see e.g. Stern et al., 1986a; Spapen, 2003; Derijke and Uitzinger, 2004; Black et al., 1985; Lutzenhiser, 1993; Bais et al., 1994; Leidelmeijer and Grieken, 2005). In addition, a contextual factor representing the reliability of vendor also plays a role in taking purchase decision (see Stern et al., 1986a; Görts et al., 2002).
Within this analytical approach, it is possible to identify a number of determinants that are evaluated as not so important(or the least significant)for household’s investment behaviour: social contacts (Derijke, E. et al., 2001); positive experience (Bais, J.M. et al., 1994; Derijke, E., Uitzinger, J., 2004); age, family age, becoming of age (Leidelmeijer, K. and van Grieken, P., 2005; Abrahamse, W. et al., 2005). Based also on a literature review, Uitdenbogerd et al. (2007:1849-1850) defined the following determinants affecting energy investments-behaviour, namely: knowledge, information about costs, information about additional characteristic (such as comfort, appearance, status, visibility and luxury), availability of choice possibilities, financial space and the ‘right’ moment.
2.3Innovation-related studies
LaBay and Kinnear (1981) looked into framework for adoption and diffusion of innovations. The authors focused on solar energy technologies as major technological innovations. Their model included both adopters and non-adopters categories and was devoted particularly to the purchase decision process. Furthermore, multivariate techniques allowed calculating mean factor importance ratings among product attributes, consumers’ demographic and socio-economic characteristics (see LaBay, and Kinnear, 1981:276).
Weber et al. (1985) drew an analogy between the development of innovative housing and technological diffusion. Among the criteria they referred to were the following ones: relative advantage, risk, compatibility, complexity, and communicability. Within this context, it is found that Hall and Reed (1999) identified ‘relative advantage’ as an important attribute of innovation, which often includes energy efficiency technologies). They defined a relative advantage as a degree to which an innovation is perceived as a better alternative within number of factors. Among the latter there are such as: profits, costs, comfort, prestige, time savings, level of effort, immediacy of reward (Hall, N. and Reed, J., 1999).
2.4Implicit discount rates
There is compelling evidence that shows that households use implicitly high discount rates (e.g. up to 90% and even much higher) that hinder the adoption of efficient technologies; thus, setting greater hurdles than for conventional technologies (see Hausman, 1979; Gately, 1980; Train, 1985; Ruderman et al., 1987; Lutzenhiser, 1992; Jaffe and Stavins, 1994a, 1994b; Metcalf, 1994; Howarth and Sanstad, 1995). Depending on the income class (see below) and for the specific case of the household market behaviour, implicit discount rates have been analysed and estimated in a number of studies:
- Train (1985) found that average implicit discount rates in household purchase decisions for efficient equipments range between: i) 10 to 32% for insulation; ii) 4 to 36% for space heating, iii) 3 to 29% for air conditioning, and iv) 18 to 67% for other appliances (e.g. water heating, cooking).
- Hausman (1979) found average implicit discount rate of 25% for air conditioners (range between 9 to 39%).
- Gately (1980) estimated rather high implicit discount rates for efficient refrigerators, ranging from 45% up to 300%.
- Dubin and McFadden (1984) estimated an average discount rate of 20% for water- and space-heating measures.
- Sutherland (1991) notes that energy efficiency appliances appear to entail very high discount rates, say 50% or higher.
Within the frame of the studies mentioned above, what is of prime importance is to look at the determinants behind such high implicit discount rates identified in the reviewed literature. Although not exhaustive, various causes can explain the use of high implicit discount rates. According to a number of authors, potential causes within the household sector can be: a lack of information about cost and benefits of efficiency improvements; lack of knowledge about how to use available information; uncertainties about technical performance of investments; lack of sufficient capital to purchase efficient products (or capital market imperfections); income level; high transaction costs for obtaining reliable information; risks associated to investments; etc. (e.g. Ruderman et al., 1981; Train, 1985; Sutherland, 1991; Gates, 1993; Metcalf, 1994).In terms of some socio-economic explanations for high implicit discount rates, the ownership status is regarded as a relevant cause (Train, 1985). Hausman (1979) and Train (1985) also argue that implicit discount rates vary inversely with income class. Train (1985) argues that the relationship between low-income class and high implicit discount rates can be explained partly because low-income households have less access to capital markets and less liquid capital to invest than higher income class households. Thus, even in the presence of good information about investment returns, lower incomes households will still be unable to invest in efficient technologies if complementary economic instruments are not in place.