Académique Documents
Professionnel Documents
Culture Documents
169
ANRV325-EG32-06 ARI 14 September 2007 20:16
Contents
1. INTRODUCTION . . . . . . . . . . . . . . 170 SOCIAL AND
Organization . . . . . . . . . . . . . . . . . . . . . 171 ENVIRONMENTAL
2. THE DECISION CONTEXT: PSYCHOLOGY . . . . . . . . . . . . . . . . . 181
RESIDENTIAL ENERGY USE Information and Incentives . . . . . . . 181
AND THE ENERGY Correlates of Proenvironmental
EFFICIENCY GAP . . . . . . . . . . . . . . 171 and Residential Energy
3. UTILITY-BASED DECISION Behavior . . . . . . . . . . . . . . . . . . . . . . 182
MODELS AND BEHAVIORAL Values, Attitudes, and Norms . . . . . 182
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
The theoretical emphasis of this paper is il- uence on social context on decision making.
lustrated where possible by empirical work on Each section ends with some general conclu-
residential energy supply, use, and efciency. sions on designing interventions for changing
DoI: diffusion of
This provides a common context for compar- individual behavior on the basis of the deci- innovations
ing the decision models reviewed and their sion models reviewed.
Attitude: relatively
implications for intervention design. Deci- This organization is not to label different enduring
sions also vary widely in their speed, effort, researchers or research traditions but to il- organization of an
level of conscious control, cognition (infor- lustrate the different theoretical approaches individuals beliefs
mation processing), and other factors. Intu- to individual decision making. The sections that predisposes his
or her actions toward
itive decisions are made on a very different ba- are designed to be read either sequentially or
an object, person,
sis than reasoning-based decisions (3). These independently according to the readers pref-
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
event or idea
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
different types of decisions are considered, erence and prior knowledge. In each section,
and their relevance for interventions in par- references to key texts on the relevant decision
ticular contexts is discussed. models and theories are included. Empirical
The objective of this review is fourfold: work is drawn upon selectively to illustrate the
rst, and most importantly, to review differ- main theoretical points. Energy-focused re-
ent disciplinary models and theories on indi- search on decision making and behavior varies
vidual decision making and the determinants widely across different research traditions. So-
of decisions; second, to extract general lessons cial psychologists have extensively explored
for designing interventions to inuence those residential energy behavior; so in Section 5,
decisions; third, to apply these lessons to res- frequent reference is made to existing empir-
idential energy use, particularly energy ef- ical reviews (4, 5). In contrast, Section 3 on
ciency; and fourth, to appeal for a more in- behavioral economics relies more on general
tegrated approach to research into residential lessons drawn from a theoretical body of re-
energy use and other problems of environ- search that has not directly explored energy-
mental consequence. related decisions.
A concluding Section 7 attempts to draw
together and compare the diverse disciplinary
Organization ndings and develop some more integrated
Section 2 sets out the decision context of resi- conclusions on intervention design and fu-
dential energy use, with reference to the en- ture behavioral research on decision making
ergy efciency gap. The body of this review and residential energy use. Three themes that
then comprises sections on decision models emerge repeatedly in the review are discussed:
from different social science traditions, orga- the importance of matching decision models
nized along a rough continuum from those to decision contexts, the scale over which in-
with an individual framing of decision making uences on decisions act, and heterogeneity
to those with a social framing. Section 3 con- in decision making.
siders economics and behavioral economics,
with a brief reference to consumer psychol-
ogy and marketing. Section 4 reviews technol- 2. THE DECISION CONTEXT:
ogy adoption theory, focusing on the diffusion RESIDENTIAL ENERGY USE
of innovations (DoI) tradition, and attitude- AND THE ENERGY EFFICIENCY
based decision models. Section 5 describes de- GAP
cision models that are empirically based in Energy efciency provides an interesting
energy-related behavioral research from so- empirical context for considering individual
cial and environmental psychology. Section 6 decision making and behavior owing to the
covers sociological questions on the agency of persistence of a gap between technological
the individual as decision maker and the in- and economic potential, and actual market
behavior (6). This energy efciency gap is re- tics (simple decision rules) on the outcomes of
peatedly identied in economic-engineering decisions.
studies that quantify the potential reduction in
Heuristics: simple
decision rules energy demand (or greenhouse gas emissions)
from the adoption of different technologies Utility Maximization and Rationality
Utility: a construct
in economics that (see Reference 7 for a review). Life cycle Microeconomic theories of consumer choice
measures an cost analysis shows short payback periods for are based on the assumption that individuals
individuals expressed the required capital investment in these tech- seek to maximize utility given budget con-
preferences for nologies (8), yet they remain underutilized. straints. A decision outcome with higher util-
different decision
Explanations for the energy efciency gap in- ity will be consistently preferred to an alter-
alternatives
clude a lack of relevant information on avail- native outcome with lower utility. Utility is
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
able technologies, limited access to capital, a construct that measures the preferences ex-
misaligned incentives, imperfect markets for pressed for different outcomes (15), but it is
energy efciency, and organizational barriers often regarded as a proxy for well-being, per-
(9, 10). sonal benet, or the betterness of an out-
Many of these market and nonmarket fail- come (16).
ures relate to individual decision making and Utility theory is derived from axioms of
indeed are pervasive facets of human behavior preference that provide criteria for the ra-
(4). These include (a) aversion to risk, uncer- tionality of choice (17). Consumers are as-
tainty, and irreversibility; (b) use of high short- sumed to behave as rational actors in a nor-
term discount rates; (c) heterogeneity of pref- mative sense of having preferences that are
erences within a population; (d ) transaction ordered, known, invariant, and consistent.
costs of searching for and processing informa- Utility-based decisions are guided by an indi-
tion; (e) sensitivity to changes in the attributes viduals evaluation of outcomes and so are es-
of energy services; and ( f ) the relative unim- sentially instrumental and self-interested (2).
portance of energy costs as a proportion of However, the rational actor model can in-
total expenditure (912). It is widely accepted corporate utility from many different sources
that interventions to reduce the energy ef- (other than money), including the perceived
ciency gap need to address these and other fairness of the decision process itself (18).
behavioral factors (13). The decision models
reviewed in Sections 36 offer many sugges-
tions in this regard and help identify reasons Utility Theory and Residential
for the varying success of interventions to pro- Energy Decisions
mote energy efciency (see Reference 14 for Utility theory and rational choice provide the
a review). building blocks for a broad range of eco-
nomic theory and practice (19). Two applica-
tions with relevance to residential energy use
3. UTILITY-BASED DECISION are discrete choice modeling and economic-
MODELS AND BEHAVIORAL engineering analyses.
ECONOMICS Discrete or qualitative choice models rep-
This section briey describes the microeco- resent individuals choices between different
nomic decision model of utility maximization alternatives characterized by a number of at-
given xed preferences. Discrete choice appli- tributes. Choices can be made as stated pref-
cations of this rational actor model in a res- erences through survey instruments or as re-
idential energy context are discussed. Utility vealed preferences through actual purchasing
theory also provides the framework for a more behavior (see Reference 20). In a residential
extensive review of behavioral economic nd- energy context, discrete choice methods have
ings on the inuence of context and heuris- been used to estimate individuals discount
rates as revealed or implied by their prefer- geneity across both decision makers and deci-
ences for energy efcient appliances (8, 21). sion contexts as well as their dynamic nature
Discount rates measure an individuals will- (30). The poor characterization of heteroge-
ingness to exchange present consumption for neous preferences is one reason why macroe-
future consumption, for example, by spending conomic models can fail to capture the energy
more up front on an appliance with lower op- efciency gap (31).
erating (energy) costs. An important nding Econometric models can successfully de-
is that individuals use different discount rates scribe behavior at an aggregate level because,
for different types of goods in different con- although individual preferences are highly
texts (22). In the case of domestic energy tech- variable, the distribution of preferences across
nologies, revealed discount rates were found a population is relatively stable (32). Modeling
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
to be clustered in the 5% to 40% range, but many individuals, in effect, assumes represen-
higher rates were applied to refrigerators and tative decision makers who may not be utility
water heaters than to heating equipment and maximizers but who, considered in aggregate,
weatherization measures (21). Other studies behave as if they were (33).
have found short-term discount rates as high
as 300% for air-conditioning technologies
(23). This marked variability suggests that dis- Irrationality and Behavioral
count rates are inuenced by many elements Economics
of the decision context, including perceived Behavioral economists seek to integrate a
risk, framing, and social arrangements (22). more robust psychological understanding of
(See sections below for additional discussion.) decision making into microeconomics. As
Discrete choice methods have also been noted above, utility theory and its applica-
used to assess the effectiveness of different tions rest on axioms of preference that broadly
types of nancial incentive in inuencing con- dene rational choice. However, there is a
sumers preference for high-efciency appli- wealth of experimental and eld evidence
ances (24). Results suggested that loans would showing that individuals do not make con-
have a larger impact than rebates in the case of sistently rational decisions (34). Time incon-
refrigerators and at a lower unit cost to the en- sistency, framing, reference dependence, and
ergy utility. Other recent examples of discrete bounded rationality are all examples discussed
choice modeling of energy-related behavior below. In each case, individual choices violate
include customers choice of electricity sup- one or more of the axioms of preference on
plier in deregulated markets (25). which utility theory is based and so are irra-
Engineering-economic analyses at an ag- tional in normative terms.
gregated sectoral or market scale (7) also im-
ply a rational actor whose preferences for dif-
ferent energy technologies are guided by a Time Inconsistency
monetary cost-benet analysis using a con- Having time-consistent preferences means
stant discount rate. Nonnancial costs or ben- that a decision taken to maximize utility from
ets can be monetized using contingent valua- a stream of both current and future values
tion or other approaches at an aggregate level will remain optimal in the future (35). Time
(26) and at the level of a residential energy user consistency is ensured by trading off present
(27). Additional transaction costs or intangi- for future consumption at a constant discount
ble costs can be included to capture the be- rate (36). Consumers purchasing behavior of
havioral factors (described in Section 2) that energy efcient appliances is used to reveal
contribute to the energy efciency gap (28, these discount rates (see above). From a be-
29). However, this monetization of transac- havioral perspective, however, this method-
tion costs can fail to recognize their hetero- ology is highly problematic (37). Extensive
empirical and experimental evidence reveals initial anchor point. More generally, the sta-
that individuals do not make decisions in a tus quo or default option of a decision tends
time-consistent manner using a constant dis- to be favored (43).
count rate (22). An immediacy effect gives Loss aversion, anchoring, and status quo
rise to high short-term discount rates when bias all illustrate the importance of fram-
otherwise immediate consumption is delayed, ing and context on decision making. There
but this is accompanied by a decline in dis- are two key implications for the microeco-
count rates over the longer term (36). The nomic decision model: (a) utility is depen-
implication is that when all costs and benets dent on a reference point; and (b) utility is
are in the future, individuals are farsighted in carried by gains and losses relative to this ref-
their advance planning, but when some costs erence point, not nal outcomes (44). Refer-
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
or benets are immediate, decisions will be ence dependence also explains why expecta-
very shortsighted (34). Consumers can also tions about decision outcomes are important
be aware that their impulse to consume now (18). If expectations set a high reference point,
impacts negatively on their long-term self- certain outcomes may be perceived as losses or
interest (23). as unfair, reducing the utility associated with
Hyperbolic or proportional discount func- consumption (45, 46).
tions have been proposed as more accurate At the household level, framing and refer-
representations of how individuals value costs ence dependence mean that income and bud-
and benets over time (23, 38). Yet noncon- geting decisions may be assigned to different
stant discount rates are rarely incorporated mental accounts (47, 48). As an example,
into utility-based decision models. One rea- an individuals willingness to spend earned in-
son is that time inconsistency makes equilib- come, windfall income, and saved income is
rium models intractable, as do other violations rarely the same even though the money in
of utility theorys axioms of preference (39). each case is fully interchangeable (49). Con-
sumption that is apparently suboptimal ac-
cording to utility theory can be explained by
Framing and Reference Dependence differences in the decision criteria used in dif-
Framing effects show that individual prefer- ferent mental accounts (50).
ences are not xed or invariant. A decision
frame refers to all the different elements that
comprise a decision: alternatives, attributes, Bounded Rationality and Decision
outcomes, and probabilities (40). The way Heuristics
these elements are presented to the decision To maximize utility within budget constraints
maker can inuence the decision outcome. requires rational actors to acquire, analyze,
Simply by framing one decision as a choice and trade off information about all possible
between losses and another as a choice be- alternatives before making a decision. Early
tween gains, preferences can be reversed even research found that the way information is
though the outcomes and their expected val- structured in different decision contexts can
ues are identical in both decision contexts inuence choices (51). In other words, prefer-
(17). This is because individuals are generally ences are inuenced by the cognitive burden
averse to losses. of information gathering and processing. In-
When making a decision, individuals also dividuals rationality (in a normative sense) is
tend to anchor on certain types of infor- bounded by these psychological and environ-
mation, rather than search for and process mental constraints (33).
all relevant information (41, 42). Depending Rather than always seeking to maximize
on how information is presented or otherwise utility, decision makers use a wide range of
available, preferences can be biased toward the rules or heuristics to help reduce cognitive or
two most expensive alternatives) (54, 55). conditions (58) and are widely applied in con-
Another class of heuristics explains how sumer psychology and marketing. Marketers
decision makers retrieve information (both in- routinely exploit information gathering and
ternally and from external sources) to help retrieval heuristics such as anchoring (59).
characterize decision alternatives. The avail- A common example is for manufacturers
ability heuristic means that this retrieval pro- to add inferior options to a product range
cess can be biased in favor of information that to increase consumers preference for the
is readily available (e.g., very personal, re- superior (and more expensive) alternative
cent, or repeated) or particularly salient (e.g., (60). By setting appropriate reference points,
vivid, atypical, or otherwise memorable). Al- marketers also look to inuence decision
though more easily recalled, such informa- makers expectations. Where expectations are
tion may not represent typical conditions and, embedded in routines (see Section 6), inter-
when combined with anchoring (see above), ventions to change behavior must be framed
can bias decisions. Emotions are also an im- appropriately to avoid loss aversion (which is
portant heuristic, particularly when assessing why energy efciency is a preferred framing
risk (56). As an example, an emotional reac- over curtailment). Much consumption results
tion to an alternative can substitute for other not from rational deliberation but from
attributes that do not come readily to mind (3). automated cognitive or affective responses to
Heuristics allow cognitive effort to be stimuli (61), so the establishment and consoli-
matched to the particular structure of a deci- dation of habits, routines, mental associations,
sion (55). Given their potential inuence over and emotional reactions are key for inuenc-
decision outcomes, how and why particular ing the selection of decision heuristics (62).
heuristics are selected in different decision Emphasizing one particularly salient or emo-
contexts are important research questions. tional attribute may inuence a decision more
Detectable analogies, past cases, or exemplars than providing information on all attributes.
with a high degree of correspondence com- Selling comfort and fullled desires can moti-
monly inuence heuristic selection (32, 57). vate homeowners to renovate their home bet-
As yet, however, there is no generalized the- ter than the prospect of energy efciency (63).
ory, so empirical evidence is the best guide (3). Behavioral decision researchers combine
the normative analysis of utility theory with
the psychological insights of behavioral eco-
Lessons for Interventions nomics to support better decision making
The central implications of the rational actor (64). (See the Behavioral Decision Research
model for interventions are to improve the in- sidebar for additional discussion.) Decision-
strumental outcomes (i.e., net benets) of the structuring tools are used to clarify informa-
desirable alternative and to ensure sufcient tion, mitigate biases, and help an individual
information is available for reasoning-based understand his or her own interests and values
Behavioral decision research provides an alternative tra- electricity supplier in deregulated retail mar-
dition within cognitive psychology to this psychophysical ap- kets (68). Interventions to emphasize habit-
proach (192). Testing the normative assumptions of utility the- ual behavior and remove the risk of switch-
ory in real-world decision making reveals mechanisms that are ing supplier might encourage households to
contingent, context dependent, time sensitive, reactive, and it- make better nondefault choices (69). Alterna-
erative. Preferences based on an individuals values are often tively, default choices (or behaviors) can be
constructed as they are being elicited (193). As values (and dened in the public interest as is common
beliefs, attitudes, or norms) are difcult to measure without in other public policy domains, e.g., health
asking, specifying psychological decision models is more an ar- and crime (70). For example, removing default
chitectural process than an archaeological one (194). Decision temperature settings from washing machines
researchers in this tradition emphasize the preference elicita- was found to reduce energy usage by 24% as
tion process, prior knowledge or mental models, biases and users set lower washing temperatures using
cognitive limitations, as well as the need for a normative analy- the new implicit anchor point of zero (71).
sis from the decision makers perspective. Structured methods Appropriate anchor points on Energy Star or
can then be used by decision analysts to help individuals make other energy efciency product labels might
better decisions (15, 40). Lessons for interventions are simi- also be inuential.
larly prescriptive but are focused on the decision process. By Mental accounting provides an alternative
comparison, the lessons for interventions considered in this framework for assessing energy efcient ren-
review concentrate on (socially desirable) decision outcomes. ovations or weatherization. In conventional
economic analyses, payback periods for capital
investments are calculated from energy sav-
(15, 40). However, these types of intervention ings. In making decisions, however, individu-
are focused on the decision-making process als may partition monetary/nonmonetary, en-
and tend to remain neutral as to the decision ergy/nonenergy, positive/negative elements
outcome. into different mental accounts and assess them
Outside marketing therefore, behavioral separately. Where minimally satisfactory out-
economic ndings are not typically used to de- comes in each account cannot be achieved,
sign interventions for changing behavior. An individuals might search for new alternatives
exception is found in the case made in legal (53). Interventions cannot therefore simply
journals for paternalistic regulations to help rely on short-term monetary paybacks if there
consumers avoid suboptimal choices that do are perceived losses in comfort or service qual-
not serve their own interests insofar as they ity. Net benets may be needed in each mental
would be made differently with complete in- account, not just overall, particularly if there
Norm: an expected formation, coherent and ordered preferences, are salient opportunity costs to investment
pattern of behavior, and unlimited cognitive resources (65, 66). In (e.g., if renovate and go on holiday are alter-
either social or
personal
this vein, one study of refrigerator purchases natives for expenditure in the saved income
demonstrated that efciency standards had account).
Internal feedback:
reinforce attitudes,
ARI
resolve dissonance
178
Cognitive: Affective: Dissonance? Embed
Wilson
awareness/ attitude/ intention/ Behavior change/
14 September 2007
Dowlatabadi
Innovativeness Attractiveness
of adopter of innovation
Characteristics Perceived attributes
of decision maker: of the innovation: External feedback:
socioeconomic status relative advantage diffusion through
personality variables compatibility social networks and
communication behavior complexity Reduce other communication
trialability uncertainty channels
observability
Selective exposure/perception
Change
Prior conditions:
practices/
perceived need or problem
norms
social norms, behavior
previous and existing practice
Figure 1
The innovation decision process. Adapted with the permission of The Free Press from Diffusion of Innovations, fth edition, by Everett M. Rogers, copyright
c 2003 (73).
ANRV325-EG32-06 ARI 14 September 2007 20:16
Table 1 Attributes of innovations that support adoption decisions, with examples from studies of residential energy use
Attribute (from DoI) and its description (73) Example of attribute in a residential energy context
Relative advantage over the incumbent technology or Cost savings, personal comfort, and family health from
practice (e.g., more convenient, exible, cheap) weatherization measures (76, 77)
Compatibility with existing needs or problems, prevailing Energy efciency is unattractive if framed as a major deviation
social norms, and behavior from behavioral norms (78)
Complexity, i.e., the skills, capacity, and effort required to A perceived barrier to solar photovoltaic adoption (79)
adopt an innovation
Trialability, e.g., whether innovations can be tested prior to Peer experience or social feedback is important to reduce
adoption uncertainty about irreversible weatherization measures (76);
conversely, clock thermostats can be tested in situ (80)
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
Observability, e.g., whether innovations are highly visible Solar technologies have greater normative appeal than less visible
(to potential adopters) measures such as home insulation (81)
Other behavioral models linked to DoI tage) and solar photovoltaics (observable,
shift the focus from psychological processes trialable/reversible) have complementary
to social communication and feedback. The attributes that may aid adoption if packaged
hierarchy of effects model examines how dif- together as a zero-emission home or its
ferent communication channels (e.g., mass equivalent. Second, barriers to adoption
media, or person to person) inuence decision should be positively identied (rather than
making at each layer of a hierarchy from in- assumed to be the inverse of drivers of adop-
formation, knowledge, attitude, and intention tion), and interventions should be designed
through to behavior (97). For example, mass to overcome them (85). Third, DoI clearly
media channels were found to inuence gen- distinguishes between the different types of
eral energy awareness but have little impact adopter (see the characteristics of decision
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
information (106, 107). Homeowners of zero- Residential energy users were found to be
emission homes in a California residential consumers of intangibles, members of social
development communicated the energy cost groups, committed individuals, and problem
Proenvironmental
savings to neighbors in standard efciency avoiders as well as informed economic ratio- behavior:
homes (108). This type of social feedback on nalists (112). Although monetary incentives intentional behavior
outcomes is important in supporting positive certainly have a calculable effect on monetary with a reduced
attitude formation (see Figure 1) and in cost-benet ratios, their impact on decisions environmental
impact relative to
complementing normative beliefs (84). Social are more contingent. For example, the attrac-
comparable
marketing approaches aim to embed behavior tiveness of incentives varies across different behaviors
change in a social context through public target groups. Administrative effort, eligibil-
commitments monitored by the participating ity criteria, cash-ow timing, the relevance of
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
community (109). Establishing social norms immediacy (see Section 3 on discount rates),
works most effectively for technologies or and the requirement to take on debt are all
behaviors that are observable by potential attributes of an incentive, which may affect
adopters (110), favoring solar photovoltaics its uptake and bring social and psychological
over insulation for example. Interventions at factors into play (113). Similar ndings on the
the community level are particularly relevant effectiveness of incentives apply to other types
where social norms at the household level of proenvironmental behavior (114).
might actually be barriers to adoption as in Providing information to raise awareness,
some cases with photovoltaics (79). Social substantiate beliefs, and inuence behavior
feedback, communication, and reinforcement was also not found to be universally effective
approaches (see the hierarchy of effects and (see Reference 5 for a detailed review).
precede-proceed models) are used in other Whereas diffusion studies explored the inu-
domains but should undergo additional ence of different channels of communication,
testing in a residential energy use context. social psychologists focused more on the form
and content of information. A key set of nd-
ings was that the most effective information
5. DECISION MODELS IN in promoting residential energy efciency
SOCIAL AND ENVIRONMENTAL was simple, salient, personally relevant, and
PSYCHOLOGY easily comparable rather than technical,
Social and environmental psychologists have detailed, factual, and comprehensive (115,
focused extensively on residential energy ef- 116). The perceived trustworthiness and
ciency. Early research from the 1970s ex- credibility of the information and/or service
plored the inuence of information and incen- provider was also important (117, 118).
tives on residential energy use behavior (111). As in DoI (see Section 4), social psychol-
As energy prices and the associated incentives ogists emphasized the importance of infor-
to conserve ebbed through the 1980s, atten- mation as feedback (119, 120). To conserve
tion shifted from decient information as a energy, home occupants must know how be-
source of market failure to the role of psycho- havior and energy use interrelate and must
logical constructs (values, attitudes, norms) be motivated to conserve (121). In this sim-
framed by environmental concerns. plied model, information provides the for-
mer, incentives provide the latter, but only
feedback provides both. Periodic and undif-
Information and Incentives ferentiated utility bills hinder direct feedback
Social psychologists demonstrated the short- on energy efcient behavior (122). Alterna-
comings of the rational actor models undiffer- tive mechanisms for conrming cost savings
entiated focus on information and net mone- include visible consumption meters, differen-
tary benets as determinants of decisions (78). tiated billing, and smoke sticks (which indicate
leaky building envelopes) (96, 123). Targeted, repeatedly reinforcing curtailment behaviors,
personalized, or otherwise tailored informa- such as thermostat setting (134). Household
tion is also important (124). Web-based tools characteristics (age, size, state of repair), com-
VBN:
value-belief-norm can now combine large sample sizes with tai- position, and occupancy are also important
(theory) lored information (125). Electricity market for heating and comfort-related measures (5,
deregulation has enabled further experimen- 132). A review of more than 40 U.S.-based
tation with time-of-use or real-time pricing to studies of residential energy use found that
support feedback (see Reference 126 for a re- while attitudes correlate with intentions to
cent review). As a recent example, load reduc- change behavior, house characteristics better
tions of 25% and 13% for California house- predict actual actions like weatherization
holds with and without automated response (135).
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
Values
Market
Threat of adverse actors
consequences to values Socioeconomic
14 September 2007
status
Technologies
Assignment of responsibility to self
Beliefs
Technical Regulations Social
20:16
Value-belief-norm theory
personal norms Supply
chain
Norms
Behavior-specific
predisposition
Habits, External
Attitudes experience Capabilities conditions
Context-specific
interaction effects
183
Activism
Figure 2
An integrated model of proenvironmental behavior. Adapted from Reference 150 (PC Stern, Towards a Coherent Theory of Environmentally
Signicant Behavior, Journal of Social Issues, with permission from Blackwell Publishing).
ANRV325-EG32-06 ARI 14 September 2007 20:16
played by values (141). As examples, altru- titudinal measures (147). External conditions
istic values may not be relevant in contexts inuence behavior both directly by dening
where individuals lack perceived self-efcacy available choices and their relative attractive-
ABC: attitude-
behavior-external (124) or where action is associated with self- ness (see Section 6) and indirectly through at-
conditions (model) sacrice or a sense of helplessness (142). The titude formation (134). In a study of curbside
relationship between values and proenviron- recycling, individuals perceptions of costs
mental behavior is discussed further in a re- in terms of time, effort, and inconvenience
cent review (143). were not found to impact behavior directly
but acted to reduce the strength of support-
ing attitudes (146). Some studies have sug-
The Importance of External gested that external conditions drive behavior
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Conditions
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
model emphasizes interaction effects between contextual domains (see Figure 2). Interven-
these different personal and contextual vari- tions seeking to change behavior need to iden-
ables (150), as well as behavior-specic char- tify and target the relevant constraints in that
acteristics (e.g., the degree of involvement, particular context (154). Because a single in-
effort, cost, time, skill requirement, nancial tervention might only inuence some of these
resources, convenience, repetition, contextual constraints (104), multiple interventions are
support, normative compliance) (136, 137). often required to act in concert (85, 155).
Multivariate analysis of the causal pathways Interaction effects between personal and
by which both personal and contextual vari- contextual variables may mean that address-
ables act on individual decisions is needed to ing one constraint causes another to emerge.
apply the model to different behaviors, and so Once an incentive creates unambiguous per-
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
generate further insights (1, 82, 151). sonal benets, interventions should target
personal variables such as the attitudes and
norms that predispose an individual to act
Lessons for Interventions rather than further increase the incentives
The social psychology research reviewed con- (113). Interventions themselves can also have
cerns individual decisions with consequences overlapping or reinforcing interaction effects.
on either residential energy use or, more A study of measures (energy tax, investment
broadly, the environment. Independent vari- subsidies, gas use regulation) promoting res-
ables that explain those decisions are specic idential energy efciency in the Netherlands
to both context and the behavior in ques- found that their combined effect on energy
tion (see Section 7). The inuence of psy- use was up to 30% less than the sum of their
chological variables is constrained by external individual effects (156). More generally, re-
conditions. As with the expected utility and search should be oriented toward interven-
attitude-based models reviewed in Sections 3 tions with the greatest potential environmen-
and 4, systemic inuences are treated as ex- tal impact rather than those of the greatest
ogenous so the timescale over which decisions theoretical or experimental interest (150).
are considered is short (e.g., weeks). The in- Different drivers of decisions change over
tegrated model shown in Figure 2 offers a varied timescales (149). Interventions de-
framework for exploring both personal and signed to address contextual variables (e.g.,
contextual variables as well as their interac- price incentives) or personal variables (e.g.,
tions. Lessons for designing interventions fol- information to reinforce favorable attitudes)
low from these basic insights. may aim for short-term change. When
Both behaviors and interventions must be behavior is strongly limited by external con-
distinguished, as well as the linkages between ditions, interventions aimed at personal vari-
them. One suggested typology identies the ables may have a weak direct impact on behav-
following linkages at a general level (85): ior but may be important in the longer term
to build political support for policy change
Moral suasion/education (change be-
and social support for norm change. The deci-
liefs/attitudes or activate norms) (153)
sion models reviewed in Sections 35 concern
Policy/regulation (change external con-
behavior change over shorter time frames,
ditions/incentive structure)
but the sociological models considered in
Supply chain (changes decision alterna-
Section 6 provide more insight into longer-
tives or their relative attractiveness)
term dynamics. For ongoing or repetitive be-
Community management (changes so- haviors (e.g., thermostat setting), the variables
cial norms) (109) that inuence behavior will change over time
Any given behavior will typically have as knowledge replaces preconception, bene-
many constraints across both personal and ts and costs become clearly dened, habits
and routines are established, and social norms that enable normal and socially acceptable
cede to personal norms (148). activities to be carried out as part of routine
Finally, information can support posi- domestic life (158). The demand for energy
tive attitude formation and can potentially is therefore indirect, created by services such
reinforce or inuence beliefs that activate as comfort and cleanliness, which are in
values and so create personal norms (as turn provided by devices (e.g., light xtures,
recognized by marketers). Information and washing machines) and by infrastructures
education-based interventions from multiple (e.g., transmission grids) (159). Demand is
sources or channels should be targeted, per- not a consequence of individual decisions or
sonalized, timed to take advantage of windows beliefs manifest over short time frames but is
of opportunity, and combined with other in- something that is systematically congured
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
terventions particularly where external condi- over the long term (160). Accordingly, it is
tions are strongly limiting (82). unsurprising that individual decision models
centered on psychological variables can
6. THE SOCIAL only weakly explain energy use (161). The
CONSTRUCTION OF DECISION broad empirical support for models such
MAKING as DoI (see Figure 1) relies not on the ex-
planatory power of its underlying individual
The literature reviewed in this section consid-
decision model but on uniformity in the
ers the broader social and technological con-
social and technological context for diffusion
text in which residential energy use decisions
(162).
are embedded. This sociological orientation
questions the relevance of individual decision
models and shifts the emphasis from energy Revisiting the Energy Efficiency Gap:
using behavior to the role of, and demand for, Embedded Energy Use
energy services. The conventional framing of the energy ef-
ciency gap (see Section 2) allows the prob-
Energy Demand as a Social lem to be dened technically and resolved by
Construct targeting individuals with universally appli-
Although the preceding sections in this cable technologies, practices, and standards
review have considered individual behavior (162). The sociotechnical perspective, how-
from different perspectives, they have shared ever, contends that household uses of energy
the basic assumption that the individual is an technologies are adaptive responses to partic-
autonomous decision maker, albeit subject to ular local conditions and norms and thus are
external inuences. Consequently, the sub- highly heterogeneous (159, 163). The social
jects of decision models are individuals. This dimension of residential energy use is needed
assumption is contested by researchers who to understand the energy efciency gap. Four
argue that individual decisions are instead key characteristics (161) are
constructed or determined by social and Embeddedness (home life is inherently
technological systems. Needs, attitudes, and energetic; habitual activities such as
expectations are not individual in nature but cooking, cleanliness, child care, mobil-
are part of a complex relationship between ity, and entertaining all embed energy
social norms and relations, technologies, consumption into daily routine);
infrastructures, and institutions (157). Constraints on choice (the supply-chain
From this sociotechnical perspective, of technologies, an individuals skills
individuals do not make decisions to consume and knowledge, and the disposition of
energy or the resources that provide energy. tradesmen and contractors, all constrain
Rather, energy provides useful services individual choice);
Counter marketing (the dominant mes- ity to control and customize the indoor ther-
sage to which households are exposed is mal environment. Space conditioning (heat-
not conservation but consumption); and ing and cooling) now accounts for around
Sociotechnical
Impetus (because oil shock-driven en- 50% of U.S. household energy use (166). regime: the
ergy prices have receded in memory, en- The embeddedness of energy use in do- structured web of
ergy efciency as an issue has lacked a mestic routines is reinforced by the counter interrelationships
systematic driving force). marketing of newly available and desirable between social
norms, human
energy devices and the services they provide
behavior, and
Some of these points are incorporated in (e.g., under-oor heating to warm cold tile technological
the individual decision models reviewed in oors in air-conditioned homes). The market- systems
earlier sections. Impetus relates to the issue of ing strategies used to sell these services indi-
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
salience (Section 3) and to the importance of cate the myriad social roles played by energy
both the form and content of information pro- technologies: display, status, self-expression,
vided to households on energy efciency (Sec- conventionality, convenience, security, inde-
tion 5). Constraints on choice are (or can be) pendence, and exibility (167, 168).
represented through the choice sets available
to individuals in different decision contexts
(Section 3), through perceived behavioral Analyzing the Social Organization of
control or self-efcacy variables (Section 4), Residential Energy Use
and through capabilities (Section 5). The embedded social dimension of energy use
The real wedge between individual de- is organized at different levels or scales. Anal-
cision models and this social dimension of ysis at the household (rather than individual)
energy use is embeddedness (159). The use level captures normative and routine behavior
of space conditioning technologies to make and recognizes the specialization of domes-
homes comfortable is a good example. House- tic roles (169). The number, age, gender, and
holds needs and expectations for thermal income of household occupants can be used
comfort have evolved over time. So too have to create meaningful sociocultural units: Two
the design of houses (e.g., room sizes, win- wealthy retirees have energy needs that differ
dow area), energy technologies (e.g., furnaces, from a low-income family with three young
thermostats), supporting infrastructure and children (170). The household as a unit of
institutions (e.g., electricity grids, utility tar- analysis also allows for more anthropological
iffs, and services), as well as social norms (e.g., considerations of the role of family relations,
indoor temperatures, room occupancy pro- kinship, gender, and ethnicity on energy use
les) (158). These changes in norms and tech- (168, 171).
nologies affect one another and drive fur- Controlling for differences in building
ther change. In the three decades from 1962 design and technologies, social interactions
to 1992, use of air-conditioning (AC) spread within households give rise to fairly stable pat-
from 12% to 64% of American homes (164) terns of energy use over time and also explain
and by 2001 had risen to 75% (165). The avail- the substantial variability between households
ability and adoption of AC technologies led (see Section 5 for a review of empirical stud-
to changes in the way homes were designed: ies). These patterns are also evident at the
Verandas, eaves, thermal mass, and other level of specic energy services, such as space
means of passive cooling ceased to be inte- conditioning, washing, and bathing. The
gral features. But these changes in design had energy demand from these habitual domestic
a ratchet effect on the perceived need for AC practices has been termed the social base load
(see Reference 159 for a full discussion). This as a sociological analogy to the load analyses
changing sociotechnical regime for AC was used in energy system management (167).
part of a broader normalization of the abil- Just as energy supply systems are scaled to
meet peak demand, so too do extraordinary decisions about energy services are highly
social events and activities (e.g., having din- constrained, and individual behavior is not
ner parties or sick children) create social peak readily inuenced. Targeting psychological
loads, which determine the need for, and size variables (e.g., information/education cam-
of, service-providing technologies (e.g., large paigns to inuence attitudes) or contextual
ovens and refrigerators or heated bedrooms). variables (e.g., monetary incentives to im-
Patterned differences in household energy prove cost-benet ratios) can only achieve
demand are further explored by lifestyle stud- limited success in effecting behavioral change
ies. These take an approach similar to the in the short term (161).
market segmentation analyses used widely Although contextual variables are recog-
in marketing and DoI-based studies to ad- nized by social psychologists and diffusion re-
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
dress heterogeneity (see Sections 3 and 4). searchers as important drivers of behavior (see
Lifestyle studies extend beyond demographic Sections 4 and 5), they are also seen as mal-
variables to nd linkages between energy use leable and legitimate targets for interventions.
and proxies of broader cultural and social Sociologists, by contrast, argue that contex-
identities (172). Borrowing from life cycle tual variables are elements of highly struc-
analysis methods, indirect or embodied en- tured systems that shape, stabilize, and con-
ergy can be included as well as direct en- strain behavior and that in many cases have
ergy use (173, 174). Lifestyle is represented evolved alongside technologies over long time
by, for example, patterns of time and money periods (159). Barriers to energy efciency
allocation (175). Early lifestyle studies looked may have normative and social functions that
at both inter- and intranational linkages be- cannot simply be removed (158). Substan-
tween lifestyle, well-being, and energy use tial reductions in residential energy use can
in the context of the curtailment and con- only be achieved by developing appropriate
servation response to the oil shocks in the sociotechnical regimes. But these regimes are
1970s (176, 177). Later studies had a more ex- strongly path dependent, i.e., dened by their
plicit sociotechnical perspective by assessing historical development (179). To overcome
how changes in the structural determinants this inertia, intervention designers need to
of energy use, such as employment and hous- recognize critical moments when sociotech-
ing stock, gave rise to large-scale energy use nical regimes are openly changing and can be
trends (178). However, as with market seg- most easily inuenced (163). In the case of
mentation analysis, the underlying assump- energy utilities, for example, changing regu-
tion was that of an individual energy user mak- lation and market structure have led to service
ing decisions on the basis of external condi- differentiation and opportunities for decen-
tions (163). More fundamental questions of tralized production (160). New relationships
how and why structural changes took place in between consumer and producer are emerg-
the rst place, and how they embedded de- ing, with consequences for energy demand. As
mand for energy, remained unasked (5). an example, interruptible loads challenge the
established norm that a utility companys role
is to provide electricity to meet demand. The
Lessons for Interventions use of infrastructure is proving similarly ex-
Energy demand is deeply embedded in ible with the centralized hub-and-spoke elec-
inconspicuous norms of comfort, cleanliness, tricity network giving way to more varied and
and convenience (159) as well as throughout coexisting congurations (180). For electric-
the whole supply chain for energy services ity, these critical moments for sociotechnical
(160). This has profound implications for the regime change are apparent.
ability of interventions to impact residential Potential targets for interventions are far
energy use by changing behavior. Individuals broader than either the individual or the
energy supply chain. A range of social and lessons for designing interventions to inu-
technological systems construct energy de- ence residential energy use are restated and
mand. These include urban planning, food further developed. Directions for future re-
production, fashion, and advertising (163). In- search are also considered.
terventions should facilitate interactions be-
tween households, utilities, and all the other Comparison of Disciplinary
institutions that play some role in structuring Approaches to Individual Decision
everyday life and routine (160, 168). Rather Making
than prescribing individual behavior, the ob-
It is worth re-emphasizing that this review is
jective of these interactions should be to
far from an exhaustive review of all decision
sketch out future sociotechnical conditions
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
psychological processes but may fail to texts (150). Expectations could act as another
demonstrate real-world effect sizes. bridge between disciplines because they play a
Although they have different emphases, key role in both economic studies of decision
utility theory, behavioral economics, technol- behavior (see fairness and reference points in
ogy diffusion, attitude-based behavior, and so- Section 3) and in sociological studies of nor-
cial psychology all share the assumption that mative and routine behavior (see Section 6).
the individual can be the subject of an in- Nevertheless, appealing for integration be-
tegrated decision model comprising distinct tween models of individual behavior and mod-
explanatory variables. The main integrative els built around social and technical systems
challenge is to resolve this assumption of in- may be a quixotic simplication, idealized and
dividual agency with sociological ndings on with merit perhaps, but quite mad given its
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
sion models reviewed includes the inuence research traditions reviewed address different
of contextual factors on behavior in some way: temporal and spatial scales of decision mak-
choice sets, reference points, and framing in ing. This accounts for differences in both the-
behavioral economics (Section 3); perceived oretical preferences and empirical ndings.
behavioral control in TPB (Section 4); ca- Moving from left to right through Table 2
pabilities, regulations, economic incentives, (and sequentially through Sections 36), the
and technologies in the integrated social psy- behavioral focus moves from the short term
chology model (Section 5); and sociotechni- to the long term and from the individual to
cal regimes and the construction of demand the social. These scales can be nested within
in sociology (Section 6). In general terms, one another like Russian dolls. Behavioral
the larger the scale and/or stronger the in- economic ndings on the framing of deci-
uence of external conditions on the deci- sions can be nested within attitudinal mod-
sion maker, the more the appropriate deci- els of technology adoption. These in turn can
sion model moves through this same sequence be nested within integrated social psychol-
(Sections 36). ogy models that incorporate a broader range
Decision models also range in their func- of psychological variables, including personal
tion. Utility-based models seek to describe norms as well as contextual variables such
choices made and so preferences revealed. as social norms. The individuals that these
Psychological models seek explanations nested models describe can themselves be
for these choices at an individual or social nested within networks of social interactions
network level. Sociological models extend and within sociotechnical regimes that pro-
these explanations to the level of sociotech- vide both context and constraint. Each scale
nical regimes, which subsume individual of nesting sets boundary conditions for the ex-
choices. Distinguishing descriptive from planatory power of the decision model in the
explanatory functions is also important when scale below.
selecting and using different decision models, A similar argument has been made within
particularly when the subject of analysis is economics by distinguishing preferences used
similar. As an example, long-term elasticities in the short run to make decisions from con-
of demand in econometric models (Section 3) stitutional preferences (precommitments or
and the sociotechnical structuring of demand self-imposed rules) that constrain these de-
in sociological models (Section 6) may be cisions but are ultimately changeable in the
analogous but answer how and why questions, long run (187). Although the potential for
respectively. changing behavior increases with the nesting
In short, critical questions for inter- of more scales, the interventions required be-
vention designers to ask are where on come more systemic and long term (see Ref-
the individual-to-social, instinctive-to- erence 188 for an analogous point regarding
deliberative, psychological-to-contextual, institutional decision making).
(see Sections 3 and 4) and lifestyle studies specically to test the applicability of dif-
along dimensions of material or cultural ferent decision models to a given decision
identity (see Section 6). More work is needed context.
on segmentation methods that allow decision
making to be understood at different scales.
Nested decision models readily incorpo- A Final Appeal for Integration
rate heterogeneity because each scale suggests Resurgent interest in demand-side manage-
a source of difference, for example, in de- ment programs, framed by climate change
cision heuristics (behavioral economics), in and energy security concerns, feeds a con-
attitudes or communication behavior (DoI), crete need for applied research on energy ef-
in values or capabilities (social psychology), ciency intervention design and evaluation.
and in routines or norms (sociology). Con- A vast amount has been learned over the last
trolling for heterogeneity at one scale may three decades on human behavior and energy
help characterize better the impact of het- use. As well as reafrming this body of knowl-
erogeneity at another scale. Heterogeneity edge, this review aims to highlight the ben-
also raises important questions about equity ets of renewed and sustained collaboration
and the distributional impact of interventions between disciplinary approaches to energy ef-
(189, 190). Aggregate analyses (of the poten- ciency. Among other things, this requires
tial for, or results of, an intervention) will con- greater openness on the part of the dominant
ceal much of the detail of individual behavioral economic-engineering tradition and a more
change. This can lead to the misspecication applied focus on the part of behavioral scien-
of decision models at the individual scale (see tists. The shared goal should be to entrench
Section 3). the social and behavioral determinants of en-
Finally, heterogeneity within populations ergy use as a wholly integrated part of energy
emphasizes the benets of natural experi- efciency research.
SUMMARY POINTS
1. There are many different models of decision making and behavior within the social
sciences. These models vary widely in their basic assumptions, independent variables,
structure, and scale.
2. The determinants of decisions can be broadly grouped into psychological and con-
textual domains. Psychological determinants include values, attitudes, and personal
norms. Contextual determinants include available choices, economic incentives, social
norms, technologies, and infrastructures.
3. Most research traditions, including economics, behavioral and social psychology, and
technology diffusion, center on the individual as decision maker. In contrast, sociol-
ogists question the relevance of individually framed decision models and emphasize
the social and technological construction of behavior.
4. Decision models can inform the design of interventions to change behavior by iden-
tifying the key inuences on decision making. Each of the research traditions offers
specic lessons for intervention design.
5. When applying behavioral theories to the design or evaluation of an intervention, an
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
appropriate decision model must be selected to match the particular decision charac-
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
FUTURE ISSUES
1. Various integrated approaches to decision making have been proposed, but
these approaches should be developed further and tested in different decision
contexts.
2. Researchers should develop general rules or guidelines to help intervention design-
ers select decision models that are appropriate for understanding how to inuence
particular behaviors.
3. New frameworks and methods are needed to reconcile individual scale (and predomi-
nantly quantitative) decision models with social scale (and predominantly qualitative)
representations of behavior.
4. Better methods for characterizing heterogeneity in decision making and behavior
across the target population for an intervention are needed.
DISCLOSURE STATEMENT
The authors are not aware of any biases that might be perceived as affecting the objectivity of
this review.
ACKNOWLEDGMENTS
We are very grateful to Baruch Fischhoff, Paul Stern, and Ashok Gadgil for helpful advice
and comments. This review was supported by the Climate Decision Making Center created
LITERATURE CITED
1. Brewer G, Stern PC. 2005. Decision-making for the Environment: Social and Behavioral
Science Research Priorities. Washington, DC: Natl. Acad.
2. Jackson T. 2005. Motivating sustainable consumption: A review of evidence on con-
sumer behaviour and behavioural change. Rep. to Sustain. Dev. Res. Netw., Univ. Surrey,
Guildford, UK
3. Kahneman D. 2003. Maps of bounded rationality: psychology for behavioral economics.
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
4. Stern PC. 1992. What psychology knows about energy conservation. Am. Psychol.
47:122432
5. Lutzenhiser L. 1993. Social and behavioral aspects of energy use. Annu. Rev. Energy
Environ. 18:24789
6. Jaffe AB, Stavins RN. 1994. The energy efciency gap: What does it mean? Energy Policy
22:80410
7. Brown MA, Levine MD, Romm JP, Rosenfeld AH, Koomey JG. 1998. Engineering-
economic studies of energy technologies to reduce greenhouse gas emissions: opportu-
nities and challenges. Annu. Rev. Energy Environ. 23:287385
8. Ruderman H, Levine MD, McMahon JE. 1987. The behavior of the market for energy
efciency in residential appliances including heating and cooling equipment. Energy J.
8:10124
9. Brown MA. 2001. Market failures and barriers as a basis for clean energy policies. Energy
Policy 29:1197207
10. Levine MD, Koomey JG, McMahon JE, Sanstad AH, Hirst E. 1995. Energy efciency
policy and market failures. Annu. Rev. Energy Environ. 20:53555
11. Jaffe AB, Stavins RN. 1994. Energy efciency investments and public policy. Energy J.
15:4365
12. Sutherland RJ. 1991. Market barriers to energy efciency investments. Energy J. 12:1534
13. Intergov. Panel Clim. Change (IPCC). 2001. Climate Change 2001: Mitigation.
Cambridge, UK: Cambridge Univ. Press
14. Gillingham K, Newell R, Palmer K. 2006. Energy efciency policies: a retrospective
examination. Annu. Rev. Environ. Resour. 31:16192
15. Clemen RT, Reilly T. 2001. Making Hard Decisions. Pacic Grove, CA: Duxbury
16. Kahneman D, Diener E, Schwartz N. 1999. Well-being: The Foundations of Hedonic
Psychology. New York: Russell Sage
17. Tversky A, Kahneman D. 1981. The framing of decisions and the psychology of choice.
Science 211:45358
18. Thaler R. 1985. Mental accounting and consumer choice. Mark. Sci. 4:199214
19. Starmer C. 2000. Developments in nonexpected utility theory: the hunt for a descriptive
theory of choice under risk. J. Econ. Lit. 38:33282
20. Louviere JJ, Hensher DA, Swait JD. 2000. Stated Choice Methods: Analysis and Application.
Cambridge, UK: Cambridge Univ. Press
21. Train K. 1985. Discount rates in consumers energy-related decisions: a review of the
literature. Energy 10:124353
22. Frederick S, Loewenstein G, ODonoghue T. 2002. Time discounting and time prefer-
ence: a critical review. J. Econ. Lit. 40:351401
28. Jaccard M, Nyboer J, Bataille C, Sadownik B. 2003. Modeling the cost of climate policy:
distinguishing between alternative cost denitions and long-run cost dynamics. Energy J.
24:4973
29. Wirl F, Orasch W. 1998. Analysis of United States utility conservation programs. Rev.
Ind. Organ. 13:46786
30. Ramesohl S. 2000. Social interactions and conditions for change in energy-related
decision-making in SMCs: an empirical socio-economic analysis. See Ref. 195, pp. 20726
31. Laitner JAS, DeCanio SJ, Peters I. 2000. Incorporating behavioural, social, and orga-
nizational phenomena in the assessment of climate change mitigation options. See Ref.
195, pp.164
32. McFadden D. 1999. Rationality for economists? J. Risk Uncertain. 19:73105
33. Conlisk J. 1996. Why bounded rationality? J. Econ. Lit. 34:669700
34. Camerer C, Loewenstein G. 2004. Behavioral economics: past, present, future. See Ref.
196, pp. 351
35. ODonoghue T, Rabin M. 1999. Doing it now or later. Am. Econ. Rev. 89:10324
36. Gintis H. 2000. Beyond homo economicus: evidence from experimental economics. Ecol.
Econ. 35:31122
37. Stern PC. 1986. Blind spots in policy analysis: What economics doesnt say about energy
use. J. Policy Anal. Manag. 5:20027
38. Harvey CM. 1994. The reasonableness of non-constant discounting. J. Public Econ. 53:31
51
39. Tversky A, Kahneman D. 1992. Advances in prospect theory: cumulative representation
of uncertainty. J. Risk Uncertain. 5:297323
40. Keeney RL. 1992. Value-Focused Thinking: A Path to Creative Decision-making. Cambridge,
MA: Harvard Univ. Press. 416 pp.
41. Tversky A, Kahneman D. 1974. Judgment under uncertainty: heuristics and biases. Science
185:112431
42. Ariely D, Loewenstein G, Prelec D. 2003. Coherent arbitrariness: stable demand curves
without stable preferences. Q. J. Econ. 118:73106
43. Baron J. 2004. Thinking and Deciding. Cambridge, UK: Cambridge Univ. Press
44. Kahneman D, Tversky A. 1979. Prospect theory: an analysis of decision under risk.
Econometrica 47:26392
45. Schwartz B. 2004. The Paradox of Choice: Why More Is Less. New York: HarperCollins.
46. Frey BS, Benz M, Stutzer A. 2004. Introducing procedural utility: not only what, but also
how matters. J. Inst. Theor. Econ. 160:377401
47. Thaler R. 1990. Anomalies: saving, fungibility and mental accounts. J. Econ. Perspect.
4:193205
48. Shefrin HM, Thaler R. 2004. Mental accounting, saving, and self-control. See Ref. 196,
pp. 395428
49. Thaler R. 1999. Mental accounting matters. J. Behav. Decis. Making 12:183206
50. Heath C, Soll JB. 1996. Mental budgeting and consumer decisions. J. Consum. Res. 23:40
52
51. Simon HA. 1956. Rational choice and the structure of environments. Psychol. Rev. 63:129
38
52. Gigerenzer G, Todd P. 1999. Simple Heuristics that Make Us Smart. Oxford: Oxford Univ.
Press
53. Simon HA. 1997. Empirically Grounded Economic Reason. Cambridge, MA: MIT Press
54. Todd PM, Gigerenzer G. 2003. Bounding rationality to the world. J. Econ. Psychol. 24:143
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
65
55. Payne JW, Bettman JR, Johnson EJ. 1993. The Adaptive Decision Maker. Cambridge, UK:
Cambridge Univ. Press
56. Finucane ML, Alhakami A, Slovic P, Johnson SM. 2000. The affect heuristic in judgments
of risks and benets. J. Behav. Decis. Making 13:117
57. Gilboa I, Schmeidler D. 1995. Case-based decision theory. Q. J. Econ. 110:60539
58. Camerer C. 2004. Prospect theory in the wild: evidence from the eld. See Ref. 196,
pp. 14861
59. Hoyer WD, MacInnis DJ. 2004. Consumer Behavior. New York: Houghton Mifin
60. Simonson I. 1993. Get closer to your customers by understanding how they make choices.
Calif. Manag. Rev. 35:6884
61. Paul Peter J, Olsen JC. 2005. Consumer Behavior and Marketing Strategy. New York:
McGraw-Hill
62. Hawkins DL, Best RJ, Coney EA. 2004. Consumer Behavior: Building Marketing Strategy.
New York: McGraw-Hill
63. Knight RL, Lutzenhiser L, Lutzenhiser S. 2006. Why comprehensive residential energy ef-
ciency retrots are undervalued. Presented at ACEEE Summer Study Energy Efc. Build.,
Asilomar, CA
64. Fischhoff B. 2005. Decision research strategies. Health Psychol. 24:S916
65. Sunstein CR, Thaler R. 2003. Libertarian paternalism is not an oxymoron. Work. Pap. 03-2,
AEI-Brookings Jt. Cent. Regul. Stud., Washington, DC
66. Camerer C, Issacharoff S, Loewenstein G, ODonoghue T, Rabin M. 2003. Regulation
for conservatives and the case for asymmetric paternalism? Univ. Pennsylvania Law Rev.
151:121154
67. Moxnes E. 2004. Estimating customer utility of energy efciency standards for refriger-
ators. J. Econ. Psychol. 25:70724
68. Brennan TJ. 2007. Consumer preference not to choose: methodological and policy im-
plications. Energy Policy 35:161627
69. Dawnay E, Shah H. 2005. Behavioural economics: seven principles for policy-makers,
New Econ. Found., London, UK. http://www.neweconomics.org/gen/uploads/
tfi0ypn1141p45zoi0mrrgf222092005201739.pdf
70. Halpern D, Bates C, Mulgan G, Aldridge S. 2004. Personal responsibility and changing
behaviour: the state of knowledge and its implications for public policy, Prime Ministers Strategy
Unit Cabinet Off., London, UK
71. McCalley LT. 2006. From motivation and cognition theories to everyday applications
and back again: the case of product-integrated information and feedback. Energy Policy
34:12937
72. Rogers EM. 2004. A prospective and retrospective look at the diffusion model. J. Health
Commun. 9:1319
73. Rogers EM. 2003. Diffusion of Innovations. New York: Free
74. Tornatzky LG, Klein KJ. 1982. Innovation characteristics and adoption-implementation:
a meta-analysis of ndings. IEEE Trans. Eng. Manag. 29:2845
75. Moore GC, Benbast I. 1991. Development of an instrument to measure the perceptions
of adopting an information technology innovation. Inf. Syst. Res. 2:192220
76. Darley JM, Beniger JR. 1981. Diffusion of energy-conserving innovations. J. Soc. Issues
37:15071
77. Brown MA. 1984. Change mechanisms in the diffusion of residential energy conservation
practices: an empirical study. Technol. Forecast. Soc. Change 25:12338
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
78. Dennis M, Soderstron EJ, Koncinski WS, Cavanaugh B. 1990. Effective dissemination
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
96. Becker LJ. 1978. The joint effect of feedback and goal-setting on performance: a eld
study of residential energy conservation. J. Appl. Psychol. 63:42833
97. McGuire WJ. 1989. Theoretical foundations of campaigns. In Public Communication Cam-
paigns, ed. RE Rice, CK Atkin, pp. 4365. Newbury Park, CA: Sage
98. Olsen ME. 1981. Consumers attitudes toward energy conservation. J. Soc. Issues 37:108
31
99. DiClemente CC, Prochaska JO. 1982. Self change and therapy change of smoking be-
havior: a comparison of processes of change in cessation and maintenance. Addict. Behav.
7:13342
100. Prochaska JO, DiClemente CC, Norcross JC. 1992. In search of how people change:
applications to addictive behaviors. Am. Psychol. 47:110214
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
101. Green LW, Kreuter M. 1999. Health Promotion Planning: An Educational and Ecological
Approach. Mountain View, CA: Mayeld. 506 pp.
102. Faiers A, Neame C. 2006. Consumer attitudes towards domestic solar power systems.
Energy Policy 34:1797806
103. Moore G. 2002. Crossing the Chasm. New York: Harper Collins
104. Egmond C, Jonkers R, Kok G. 2006. One size ts all? Policy instruments should t the
segments of target groups. Energy Policy 34:346474
105. Darby S. 2006. Social learning and public policy: lessons from an energy-conscious village.
Energy Policy 34:292940
106. Jager W. 2006. Stimulating the diffusion of photovoltaic systems: a behavioural perspec-
tive. Energy Policy 34:193543
107. Kaplan AW. 1999. From passive to active about solar electricity: innovation decision
process and photovoltaic interest generation. Technovation 19:46781
108. Hanson M, Bernstein M, Hammon R. 2006. The role of energy efciency in homebuying
decisions: results of initial focus group discussions. Presented at ACEEE Summer Study Energy
Efc. Build., Asilomar, CA
109. McKenzie-Mohr D, Smith W. 1999. Fostering Sustainable Behavior: An Introduction to
Community-Based Social Marketing. Gabriola Island, BC, Can.: New Society
110. Schultz PW. 2002. Knowledge, information, and household recycling: examining the
knowledge-decit model of behavior change. See Ref. 82, pp. 6782
111. Kempton W, Darley JM, Stern PC. 1992. Psychological research for the new energy
problems: strategies and opportunities. Am. Psychol. 47:121323
112. Stern PC. 1985. Energy Efciency in Buildings: Behavioral Issues. Washington, DC: Natl.
Acad.
113. Stern PC, Aronson E, Darley JM, Hill DH, Hirst E, et al. 1986. The effectiveness of
incentives for residential energy conservation. Eval. Rev. 10:14776
114. Katzev RD, Pardini AU. 1987. The comparative effectiveness of reward and commitment
approaches in motivating community recycling. J. Environ. Syst. 17:93113
115. Kempton W, Montgomery L. 1982. Folk quantication of energy. Energy 7:81727
116. Yates SM, Aronson E. 1983. A social psychological perspective on energy conservation
in residential buildings. Am. Psychol. 38:43544
117. Craig CS, McCann JM. 1978. Assessing communication effects on energy conservation.
J. Consum. Res. 5:8288
118. Farhar BC, Buhrmann J. 1998. Public response to residential grid-tied PV systems in
Colorado: a qualitative market assessment. Rep. NREL/TP-550-24004, Natl. Renew. En-
ergy Lab., Golden, CO
119. Katzev RD, Johnson TR. 1987. Promoting Energy Conservation. Boulder, CO: Westview
120. Farhar BC, Fitzpatrick C. 1989. Effects of feedback on residential electricity consumption:
a literature review. Rep. SERI/TR-254-3386, Solar Energy Res. Inst., Golden, CO
121. Stern PC, Gardner GT. 1981. Psychological research and energy policy. Am. Psychol.
36:32942
122. Constanzo M, Archer D, Aronson E, Pettigrew T. 1986. Energy conservation behavior:
the difcult path from information to action. Am. Psychol. 41:52128
123. Seligman C, Becker LJ, Darley JM. 1978. Behavioral approaches to residential energy
conservation. Energy Build. 1:32537
124. McMakin AH, Malone E, Lundgren RE. 2002. Motivating residents to conserve energy
without nancial incentives. Environ. Behav. 34:84863
125. Benders R, Kok R, Moll HC, Wiersma G, Noorman KJ. 2006. New approaches for
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
143. Dietz T, Fitzgerald A, Shwom R. 2005. Environmental values. Annu. Rev. Environ. Resour.
30:33572
144. Newhouse N. 1991. Implications of attitude and behavior research for environmental
conservation. J. Environ. Edu. 22:2632
145. Gatersleben B, Steg L, Vlek C. 2002. Measurement and determinants of environmentally
signicant consumer behavior. Environ. Behav. 34:33562
146. Guagnano GA, Stern PC, Dietz T. 1995. Inuences on attitude-behavior relationships:
a natural experiment with curbside recycling. Environ. Behav. 27:699718
147. Heberlein TA, Warriner GK. 1983. The inuence of price and attitude on shifting resi-
dential electricity consumption from on- to off-peak periods. J. Econ. Psychol. 4:10730
148. Olander F, Thogerson J. 1995. Understanding of consumer behaviour as a prerequisite
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
166. IEA. 2004. Oil Crises and Climate Challenges: 30 Years of Energy Use in IEA Countries, Int.
Energy Agency, Paris, Fr.
167. Wilhite H, Lutzenhiser L. 1999. Social loading and sustainable consumption. Adv.
Consum. Res. 26:28187
168. Henning A. 2005. Climate change and energy use: the role for anthropological research.
Anthropol. Today 21:813
169. Wilhite HL, Wilk R. 1987. A method for self-recording household energy-use behavior.
Energy Build. 10:7379
170. Lutzenhiser L, Lutzenhiser S. 2006. Looking at lifestyle: the impacts of American ways of
life on energy/resource demands and pollution patterns. Presented at ACEEE Summer Study
Energy Efc. Build., Asilomar, CA
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
171. Wilhite H. 2005. Why energy needs anthropology. Anthropol. Today 21:12
172. EPRI. 1990. Residential customer preference and behavior: market segmentation using
CLASSIFY. Rep. EM-5908, Electr. Power Res. Inst., Palo Alto, CA
173. Bin S, Dowlatabadi H. 2005. Consumer lifestyle approach to US energy use and the
related CO2 emissions. Energy Policy 33:197208
174. Reinders AHME, Vringer K, Blok K. 2003. The direct and indirect energy requirement
of households in the European Union. Energy Policy 31:13953
175. Weber C, Perrels A. 2000. Modelling lifestyle effects on energy demand and related
emissions. Energy Policy 28:54966
176. Nader L, Beckerman S. 1978. Energy as it relates to the quality and style of life. Annu.
Rev. Energy 3:128
177. Mazur A, Rosa E. 1974. Energy and life-style. Science 186:60710
178. Schipper L, Bartlett S, Hawk D, Vine E. 1989. Linking life-styles and energy use: a matter
of time? Annu. Rev. Energy 14:273320
179. Schipper L, Unander F, Murtishaw S, Ting M. 2001. Indicators of energy use and carbon
emissions: Explaining the energy economy link. Annu. Rev. Energy Environ. 26:4981
180. Lovins AB, Kyle Datta E, Feiler T, Rabago KR, Swisher JN, et al. 2003. Small Is Protable:
The Hidden Economic Benets of Making Electrical Resources the Right Size. Snowmass, CO:
Rocky Mountain Inst.
181. Swart RJ, Raskin P, Robinson J. 2004. The problem of the future: sustainability science
and scenario analysis. Glob. Environ. Change Part A 14:13746
182. Skumatz LA, Dickerson CA, Coates B. 2000. Non-energy benets in the residential and
nonresidential sectorsinnovative measurements and results for participant benets. Presented
at ACEEE Summer Study Energy Efc. Build., Asilomar, CA
183. Abell P. 2003. On the prospects for a unied social science: economics and sociology.
Socio-Econ. Rev. 1:126
184. Wirl F. 2000. Lessons from utility conservation programs. Energy J. 21:87108
185. Loughran DS, Kulick J. 2004. Demand-side management and energy efciency in the
united states. Energy J. 25:1943
186. Keirstead J. 2006. Evaluating the applicability of integrated domestic energy consumption
frameworks in the UK. Energy Policy 34:306577
187. Buchanan JM. 1994. Choosing what to choose. J. Inst. Theor. Econ. 150:12344
188. McDaniels TL, Dowlatabadi H, Stevens S. 2005. Multiple scales and regulatory gaps in
environmental change: the case of salmon aquaculture. Glob. Environ. Change 15:921
189. Dillman DA, Rosa EA, Dillman JJ. 1983. Lifestyle and home energy conservation in the
United States: The poor accept lifestyle cutbacks while the wealthy invest in conservation.
J. Econ. Psychol. 3:299315
190. Parker P, Rowlands I, Scott D. 2005. Who changes consumption following residential
energy evaluations? Local programs need all income groups to achieve Kyoto targets.
Local Environ. 10:17387
191. Holling CS. 1978. Adaptive Environmental Assessment and Management. London, UK:
Wiley
192. Fischhoff B. 2005. Cognitive processes in stated preference methods. In Handbook of
Environmental Economics, ed. K-G Maler, JR Vincent, pp. 93768. Amsterdam, Neth.:
Elsevier
193. Lichtenstein S, Slovic P, eds. 2006. The Construction of Preference. Cambridge, UK:
Cambridge Univ. Press
194. Gregory R, Lichtenstein S, Slovic P. 1993. Valuing environmental resources: a construc-
Access provided by b-on: Universidade Aberta (UNIV-AB) on 01/29/17. For personal use only.
Annu. Rev. Environ. Resour. 2007.32:169-203. Downloaded from www.annualreviews.org
Annual Review of
Environment
and Resources
vii
AR325-FM ARI 21 September 2007 16:39
Indexes
Errata
viii Contents