Co-evolution of legal and voluntary standards: Development of energy efficiency in Swiss residential building codes

Technological Forecasting & Social Change 87 (2014) 1–16

Contents lists available at ScienceDirect

Technological Forecasting & Social Change

Co-evolution of legal and voluntary standards: Development of energy efficiency in Swiss residential building codes Stefan N. Groesser 1 Bern University of Applied Sciences, School of Management, Strategy and Simulation Lab, Morgartenstrasse 2c, 3022 Bern, Switzerland

a r t i c l e

i n f o

Article history: Received 8 May 2012 Received in revised form 10 April 2014 Accepted 28 May 2014 Available online xxxx Keywords: Co-evolution Innovation diffusion Feedback Causal model Standard System structure Innovation ecosystem System dynamics Energy efficiency

a b s t r a c t Improving the level of energy efficiency required by building codes for refurbishments and new construction is a powerful lever for reducing greenhouse gas emissions. This paper explores how technological, social, political, and economic factors interact and shape the evolution of the energy efficiency in building codes. Existing approaches to the evolution of standards focus primarily on adopting individual or multiple technologies or products, but only peripherally explore the feedback dynamics between innovation, diffusion, and standardization (IDS.)2 To fill this void, I draw on the revelatory case of Switzerland, because the Swiss standards have continuously improved since 1970, whereas in many other countries improvements have stagnated after the recovery from peaks in energy prices. The paper's contribution is, first, a qualitative, structural model which endogenously formalizes the IDS-dynamics of standard improvement. I find that the co-evolution of voluntary and legal building codes have enabled a continuous improvement of the standards even in the absence of economic pressures. And second, I use the model for policy analysis, which indicates that several obvious policies might cause policy resistance and could result in uneconomical, counter-intuitive outcomes. Policy interventions have to dynamically balance the speed of innovation and the ability of system agents to change. © 2014 Published by Elsevier Inc.

1. Introduction Mitigating global warming and securing a mid- and longterm energy supply are relevant topics for policy makers. To limit the increase in temperature to acceptable levels, greenhouse gases (GHG)-emissions must be approximately halved by 2050 relative to 1990-levels [1–3]. The energy required for residential buildings greatly contributes to those emissions [4,5]. Therefore, improving the energy efficiency (EE3) of the residential building stock by diffusing

E-mail address: [email protected]. Tel.: +41 31 848 34 54; fax: +41 31 848 34 01. 2 Innovation–diffusion–standardization (IDS). 3 In the paper, I use EE to abbreviate both “energy efficiency” and “energy efficient”. 1

http://dx.doi.org/10.1016/j.techfore.2014.05.014 0040-1625/© 2014 Published by Elsevier Inc.

more innovative EE technologies, e.g., insulation and heating technologies for new constructions and renovations [6], are cost efficient options to lower GHG emissions [7,8]. Thereby, the improvement of the average EE of a building stock significantly depends on the energy requirements of building codes. A residential building code is a voluntary or a legal standard that defines the required level of EE (measured by the metric Energy Demand per New Constructed Housing Unit in MJ/ m2/year; see Fig. 1) in a residential building for space and water heating. History has shown that these requirements can improve over time. Innovation is the improvement in building technologies, e.g., insulation or controlling technologies [9]. Now, what causes the EE improvements in building standards? Relevant literature stems from innovation diffusion, coevolution, innovation and standardization, and technological innovation systems. For the first body of literature, Higgins et al. [10] have outlined the extensive literature on innovation

2

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

120 Building Codes (Jakob, 2008)

WDV 1982

700

Historical Data (AWEL, 2007)

MVO 1986

Energy Demand (SFOE, 2007)

Median (1975)

Energy Price (BP, 2009)

100

600 Historical Energy Demand

Median (1985)

500

80

400

60 Median (1995)

300 200

MUKen 2000

40

MVO 1992

Historical Crude Oil Price

MuKen 2008

20

100 0 1970

1975

1980

1985

1990

1995

2000

2005 Year

Cruide Oil Price (US-Dollar/Barrel)

Energy Demand per New Constructed Housing Unit (MJ/sqm/year)

800

0

Fig. 1. Historical energy demand of legal building code and oil price. The data available does not allow for precise estimates of energy demand or energy efficiency due to different measurement methods. For instance, the empirical analysis of the average energy efficiency in residential building undertaken by the Swiss Federal Office of Energy resulted in values lower than the legal building code prior to 1986, and in values higher than the legal building code thereafter [50]. The historical data available for the Statistical Office of the Canton Zurich [38] shows values significantly higher than the legal building codes [49]. For the purposes in this paper, I use the values of the building codes.

diffusion research and then used it to analyze policies for GHG-reductions in housing stocks. Although they consider voluntary and mandatory adoption of reduction technologies to evaluate the effectiveness of different intervention schemes, they assume that both technologies are independent from each other—a fact which seems to be an oversimplification as this paper will show. The second body of literature explicitly accounts for the co-evolution of innovations, e.g., supply and demand [11,12], two complementary innovations [13], clinical knowledge and technological capabilities [14], scientific and technological networks [15], or capabilities and preferences [16]. Specifically relevant is Dijk et al. [11] who provide a co-evolutionary analysis of the emergence of hybrid-electric cars. Their analysis integrates actor perspectives, feedback effects, and competition between products. The approach I use can be viewed as a more formal version or an extension of their approach. The `third body of literature deals with innovation and its standardization, specifically for processes of technology transfer and standardization [e.g., 17–19], and driving forces of standardization activities [20,21]. This paper contributes to the third body by providing a systems model with a broad model boundary which accounts for multiple agents. Thereby, it supports the most recent research on standardization cycles [19]. The fourth body of literature is on technological innovation systems. It addresses the question of how technological innovations develop [22–25] from a systems perspective. Although this literature provides insightful concepts, e.g., functions [26] or the multi-pattern approach [27], it often lacks a clear conceptualization of the process of standard development at a detailed causal level. To summarize, although the understanding of the co-evolution of EE standards is relevant from a policy perspective, it has not been directly addressed by current research.

The objective of this paper is, first, to explain the evolution of EE in building codes for the residential building sector. Innovation systems literature reveals that feedback rich models with a broad model boundary are required to adequately address such phenomena [23]. In this paper, I use the methodology of qualitative system dynamics [28,29] which, in addition to feedback dynamics, also accounts for important accumulations as well as nonlinear and delayed interactions. To understand the evolution process, I study the revelatory case of the Swiss residential building sector. The empirically grounded model4 interconnects economic, technological, and political aspects and accounts for the dynamic complexity of that system [31]. After developing a structural model based on historical case data, I use the model to discuss the likely impact of future policy interventions on the development of EE standards in building codes. I maintain that the evolution of the level of EE in a legal building code occurs in co-evolution with a voluntary building standard. Moreover, I argue that reoccurring dynamics of innovation, diffusion, and standardization (IDS) form the core of this co-evolutionary process. And finally, I argue that, due to the system's dynamic complexity, and a lack of transdisciplinary and integrated systems models for norm evolution, unbalanced policy interventions could cause policy resistance and counterintuitive outcomes. The paper is organized as follows: The research methodology is described in Section 2. Section 3 develops the case model which is then analyzed in Section 4. Section 5 uses this model to conceptualize the likely outcomes of policy interventions. Section 6 discusses the theoretical and practical contributions and implications of the results and addresses the model's

4

In the paper, I use “model” and “theory” interchangeable [30].

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

limitations. Section 7 concludes the paper and provides directions for future research. The contribution of the paper is threefold. Firstly, it contributes to research about standard development, providing an endogenous perspective on a case-specific evolution of standards. Furthermore, it shows that the co-evolution of a voluntary and a legal standard created a regime that enabled the steady improvement of EE standards even when the economic context conditions in Switzerland, i.e., low levels of energy prices, would have suggested otherwise. With respect to the endogenous perspective, the paper develops a structural feedback model that explains the co-evolution of standards development through a reoccurring innovation, diffusion, and standardization cycle. Secondly, the paper uses the endogenous model to analyze and discuss the likely impact that administrative policies could have on the future development of EE in building codes. The result of the analysis reveals that several policies may have the potential to cause unintended consequences. The third contribution is to methodology. The paper uses a qualitative systems-mapping approach known as the feedback loop method or qualitative system dynamics [32,33]. I improve the method by explicating the relevant network of stocks and flows. This is widely practiced in quantitative system dynamics, though not on the level of qualitative mapping, analysis, and design. 2. Approach and model background I applied the case-study methodology [34,35] to understand the evolution of EE in building codes used in the Swiss residential building environment. The case selected here provides a concrete illustration of the challenges of standard development and innovation diffusion in large socio-technical systems [36]. I have selected the case for theoretical, and not statistical reasons [37]. The Swiss case is particularly insightful, because the degree of EE required by the legal building code has continuously improved since 1970 [see Fig. 1, historical data: 38], whereas improvements stagnated in other countries soon after energy prices returned to “normal” values [39–42] following oil price shocks. In addition, I selected the case because I had access to the research field and continuous support from experts for a three-year period. I chose six residential buildings, with varying levels of EE, which were constructed around 2005, in the Swiss municipality of Langenthal (Canton Berne), I used this sample to trace back the related agents of the built environment [43]. This sample has a high variance regarding the buildings' EE. To ensure validity, I used multiple sources of data for triangulation, and I interviewed experts of the residential building sector throughout the research who comment on intermediate results. I drew from sociological and economic literature [44–48], numerical data [38,49–52], and from other research projects [53–55] to develop the system structures as well as to establish the development over time of important variables (e.g., the energy demand for new constructed housing units, Fig. 1) for the Swiss case. I interviewed a variety of stakeholders to collect insights from all relevant perspectives. I conducted 29 interviews, each of which lasted more than 80 min (Table 2 in the Appendix). I also conducted four collaborative workshops with the interviewed experts. The aims of the workshops were to validate intermediate

3

research results and to sample additional data about relevant system structures, interrelations, and parameters [56,57]. The collection of data and analysis interconnected during the research process, employing an iterative theory development process [30,35]. In addition, I conducted a comparative analysis of selected residential buildings to maximize the obtained insights. I formalized the analysis using causal loop diagrams and system structure diagrams [33,58], which facilitated the conceptualization process as well as the communication process with participating experts. Only the latter are reported here. A system structure diagram represents accumulations of the system under investigation and the elements that cause changes in them. Accumulations are represented as stock variables (rectangular symbol, see Fig. 2), flow variables (pipes with valves, cf. Fig. 2), which are ways to change stock variables, and intermediate variables, which are used to detail causalities. Further, the diagram shows information feedback loops, which are closed chains of causal interactions between variables [32,33,59]. Relatively recently, the consideration of feedback loops in innovation studies has received more attention [e.g., 11,26,60]. These feedback loops enable one to endogenously explain the evolution of a time series, i.e., in this case the development of EE in building codes. There are two types of loops: reinforcing loops (R) and balancing loops (B). The former accelerate initial changes in a model variable; the latter dampen such changes so that the goals or limits of a system, either implicit or explicit, are approached [33]. The diagram notation also accounts for significant time delays between cause and effect. The resulting system structure diagram is a theory and explains how the dynamic phenomenon can be created endogenously over time [30,33]. My analysis concludes with the formulation of a well-grounded theory about the co-evolution of the EE in voluntary and legal building codes. The theory consists of a set of temporally interrelated feedback loops which are derived from interviews and workshops with experts of the considered system. With this model, it is not possible to clearly derive the strength of the factors at play; it rather shows an integrative evaluation of the factors involved. The model is based on the historical development of the building code in Switzerland. I then use the model for integrative policy analysis and for developing policy recommendations. The next section details the result of the case study.

3. Resulting model about norm evolution This section develops the structural model that resulted from the case study about the historical development of the EE in residential building codes in Switzerland. In principle, I describe the structural, more abstract model, but make references to the case where they are beneficial. In the residential building sector, multiple agents are interconnected. An agent is an aggregate of individuals or organizations, e.g., private residential building owners or architects, and fulfills specific functions in that environment [43]. The next three subsections develop the structural model in successive steps. Section 4 analyses the model and relates it to the historical development of the Swiss case. In the paper, I use the terms “residential building stock” and “building stock”

4

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

Fig. 2. Dynamics in the technological and political sector (Note: A rectangle represents a stock or an accumulation, like water in a bathtub; flows, depicted as pipes with valves, fill or drain the stock. Sterman [27] provides further details about the notation. The name of the agent, to which the variable belongs, is provided in italics.)

interchangeably. All variables and loops of the model are summarized and described in the Appendix of the paper. 3.1. Technological and political sector The energy efficiency of the building stock is a physical property which changes gradually over time due to the long lifetime of residential buildings. Improving the average EE of the building stock can be attained through three actions: (1) constructing buildings with a level of EE higher than the current level of the stock, and (2) demolishing or (3) refurbishing buildings with low levels of EE. All three actions utilize the contemporary building codes (Fig. 2). In Switzerland's case, two principal standards are distinguished: a legal building code and an innovative voluntary standard. The first represents the mandatory legal norm set up by cantonal authorities [see 20 for standards from authorities] specifying threshold values of EE for residential buildings, and applies to new construction and building refurbishments. The latter standard is a voluntary building code which, in relative terms, is of higher EE than the legal standard, and hence is termed as an “innovative EE standard”. In Switzerland, both standards use the measure Energy Demand per New Constructed Housing Unit (Fig. 1) which measures the annual required MJ per square meter area in a residential building for space and water heating. In the Swiss case, the concept of EE gained recognition in the 1970s (Fig. 1). At that time, a voluntary standard began to develop; initially it was not a formalized legal code, but an implicit code of best practice which was accepted among experts such as architects, building construction companies, and craftspeople. In relation to the required EE, this implicit practice continuously improved and ultimately became explicit

and formalized, first, by the Swiss Society of Engineers and Architects (SIA) in 1980 as an SIA-norm and then in 1998 by the Minergie© society as the Minergie® standard.5 Voluntary standard as used in this paper is a standard which agents can choose to comply with. The legal standard which all agents have to comply with is under government control. This understanding is specific to Switzerland's distinct governance conditions and could be managed differently in other countries, e.g., the USA [61].6 When comparing voluntary and legal standards, one perceives a relative advantage of the innovative standard regarding EE over the legal standard. This advantage changes according to the changes made in both standards. The energy efficiency of the innovative standard is improved by making improvements in relevant residential housing technologies, e.g., insulation, heating, and controlling technologies [9]. Improvements of the legal building

5 The Swiss Society of Engineers and Architects (SIA) is Switzerland's leading professional association for construction, technology and environment specialists. The SIA and its members stand for quality and expertise in architecture and construction. The SIA is well known for its work on standards. It develops, updates and publishes numerous standards, regulations, guidelines, recommendations and documentation, which are of vital importance for the Swiss construction industry. The SIA is a member of the Swiss Association for Standardization (SNV) and thereby ensures the relation to requirements of European standards (CEN) on the topic of building construction and related materials. MINERGIE® is basically a private nonprofit organization. It is financed by its members, its services (certification, education and information programs, consulting and coaching) and its sponsors (companies of the Swiss construction industry, investors and different levels of government). 6 According to Philips [61], the USA standard development system is a truly voluntary system and is not under government control.

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

code's EE are based on the advantage in EE of the innovative standard, because the advantage demonstrate technical feasibility. The interdependence that has just been described can be formalized as balancing loop B1. It intends to improve the legal standard until the legal code reaches an EE level similar to the innovative standard. A constraining factor affecting improvements is the willingness to improve EE of legal building code among political agents. When the EE of the legal code is low, the marginal benefits of improvements in EE relative to other investment projects, e.g., fire protection, stability, and noise reduction [39], are high. In this situation, political agents are interested in these benefits and improve the legal standard. However, this willingness decreases as the EE reaches higher levels. The relative marginal benefits become smaller, which result in a lower willingness to further improve the legal code [62]. Loop B2 demonstrates that improving the EE of the legal code has limits. Next, I proceed by developing the market and social dimension. 3.2. Market sector Innovative energy efficient housing consists of accommodations that use innovative EE building standards and which have EE levels higher than the legal building code [63]. At the beginning of the development of EE in Switzerland, this innovative EE standard was an implicit best practice in the architectural industry. Even though it was implicit, it was applied to the construction of innovative EE housing. Later, the implicit innovative EE best practices became formalized and became known as Minergie®-Standard. New construction of innovative EE housing, i.e., construction that fulfills innovative EE standards, increases the stock of innovative EE housing. The aging of the technology of innovative EE housing decreases the stock accordingly (Fig. 3). To clarify, old housing still has the same EE level; what changes is that housing technology is no longer the state-of-the-art, due to technological advancements. When innovative EE housing is phased out, it becomes part of the normal EE housing. This happens when innovative housing technologies enter the market; yesterday's innovative housing technologies become today's normal EE housing technology. The same structure that has just been explained applies to the capacity to construct innovative and normal EE housing. Capacity here means the physical ability and knowledge of supply agents to construct new innovative EE housing as well as upgrade normal EE housing through refurbishment. Based on the demand–supply balance for EE housing, construction companies invest in new capacity for constructing EE housing; to ease comprehension, other investment possibilities are not detailed here. Investment in innovative EE capacity increases the supply of construction capacity for this standard of housing, which slows down further investments in this kind of capacity (B3). As innovative EE housing, this capacity can also outphase and become capacity for constructing normal EE housing. The aging of both the innovative capacity as well as innovative EE housing depends on the technology lifecycle of housing technology and the rate of technological improvement. The innovative EE housing compared to the amount of normal EE housing demonstrates the dissemination level of innovative EE concepts in the building stock [55,64].

5

Familiarity and the cost advantage of innovative concepts, such as innovative EE housing, depend on the degree of its diffusion [65–68]. Familiarity is the degree to which agents have been exposed to the innovation. In the residential building sector, this occurs mainly by the physical observability and access to trials of innovative EE housing, e.g., by demonstration and pilot programs, as well as the direct and indirect personal communication between agents [44,68–70]; thus, the agents' social exposure to innovative EE housing directly influences their familiarity [71–74]. The cost advantage of innovative EE housing for heating and warm water is favorable when comparing the costs of normal EE housing for these services. When calculating the costs, I consider only investment costs and operating costs [75]. I exclude the effects of co-benefits from the relative advantage of innovative EE housing, which will be detailed in the section on limitations. Based on the experience curve, the investment costs for innovative EE housing are reduced on average, with a doubling of the cumulative output of innovative EE housing [76,77]. The cost advantage of innovative EE housing, and the familiarity of it, results in the relative attractiveness of innovative EE housing, which further stimulates the demand for this type of housing, thereby leading to new innovative EE housing. This type of construction provides valuable experience effects, which improve the cost advantage. The physical existence of innovative EE housing also increases the familiarity with this kind of housing so that the virtuous circle R1 kicks-off and starts to fill the innovative EE housing stock. The incumbent agents who are considered next have a strong position in the residential sector due to large financial and social investments over long periods of time, e.g., assets and social relationships, which are vulnerable to technological advancements. Examples of such incumbent agents are utilities and large residential real estate owners, with their substantial investments. These agents may perceive the increase of the relative attractiveness of innovative EE housing as a threat, and might react with innovative actions themselves which aim to protect their current market position [48,78]. For instance, utilities could reduce their energy prices or improve their contracting offerings for heating technologies. These actions incrementally improve the cost efficiency of normal EE housing which, in turn, reduces the relative cost advantage of innovative EE housing (B4).7 A further action of these incumbent agents is to lobby and lower the support for innovative EE at a political level. Other agents also try to influence the political agenda. According to the experts involved in the study, utilities and large real estate owners are well established in the political lobbying system in the Canton of Bern and other cantons in Switzerland. The agents who own innovative EE construction capacity promote their innovative EE business to other contacts and political agents. This support is the degree to which agents publicly engage in enhancing the dissemination of the idea, e.g., via personal discussions, engaging in pilot or demonstration projects, or holding public meetings. The level of support accounts for the

7 The paper's model accounts for incremental innovations. Based on empirical evidence from experts of the Swiss case, the concept of radical innovations is not applicable.

6

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

Fig. 3. Dynamics in the market sector of the residential building sector.

number of activities, initiatives, or time resources dedicated to lobbying. The last concepts in Fig. 3 are technology improvement rate and lifetime of housing technology. Improvements of housing technology, e.g., better insulation, heating and monitoring equipment, shorten the lifecycle time of existing housing technology, as well as the technology for constructing EE housing. When new innovative EE technologies are developed and made available, existing innovative EE technologies are slowly phased out and become normal EE technologies, although, as explained earlier, the level of EE remains the same. As a consequence, innovative EE housing moves faster to the stock of normal EE housing. The same effect happens to the capacity for constructing innovative EE housing: it moves faster to the stock of capacity for constructing normal EE housing (Fig. 3). 3.3. Integrating market, political, and technological sectors When combining Figs. 2 and 3, the model about the evolution of the EE of voluntary and legal standards becomes rich in feedback dynamics and assumes a broader model boundary (Fig. 4). Two interface variables connect the diagrams: support for innovative energy efficiency and the relative advantage of innovative standard regarding energy efficiency. One mechanism that results from the combination is the loop B5. Assuming that additional capacity for constructing EE housing becomes available, the agent who owns this capacity lobbies for innovative EE which, with a delay, increases the EE of the legal code. This, then, lowers the technological advantage of the innovative EE standard and also reduces its

attractiveness. A lower demand for innovative EE housing results reduces new investments in this type of capacity. Thus, B5 has the potential to balance the supply of innovative EE housing. However, B5 is also highly nonlinear; political agents are only willing to improve the legal standard when a significant level of support has built up. The second loop that results from the combination is R2b. This mechanism includes the capacity for constructing normal EE housing. Owners of this type of construction capacity are, on average, opposed to improvements of the legal standards since this would force them to invest in innovative EE technologies to fulfill requirements and remain competitive. Hence, they withdraw their support and thereby impair the improvement of the legal building code's EE, with the result that the innovative standard tightens its technological advantage, and more capacity for innovative EE housing is constructed. In principle, an identical loop exists for the stock normal EE housing (R2a) which also results from the combination. The owners of this stock, on average, are also opposed to innovative EE since they would have to invest to maintain their current position. The last feedback relation in Fig. 4 is R3. This loop relates to incumbent agents (e.g., utilities) who are threatened by innovative EE housing and subsequently reduce their support for innovative EE. This action results in slower improvements in legal standards, and tightens the technological advantage of the innovative standard. To summarize, the current model (Fig. 4) accounts for the market, political, and technological sectors, and connects them by means of seven interacting feedback loops. Next, I show their likely interactions and demonstrate the use of

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

7

Fig. 4. Model with market, political, and technological sectors.

qualitative system structure diagrams [some examples: 28,79]. In Section 5, I analyze policy options. 4. Model analysis 4.1. Core dynamics of innovation, diffusion, and standardization (IDS) Assuming that the stocks of innovative EE housing and the capacity for constructing such housing are empty; hence, the stocks of normal housing and the respective construction

capacity will have high values. Further, let us suppose that all three stocks of EE have low values. This was the situation which occurred around 1970. Since 1973, public awareness of Switzerland's dependency on fossil resources has increased dramatically, with the occurrence of the oil price shocks in 1973 and 1980 [oil crises in 1973 and 1979/80: 80,81]. These shocks initiated discussions about setting environmental policy targets (variable “political target of annual energy demand” in Fig. 5) to reduce dependency on fossil resources and to increase EE [82]. The publication of the Club of Rome's study Limits to Growth [83] sensitized the general public and

8

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

Fig. 5. Policy model including the administration sector.

politicians about environmental issues. These events and the long-term vision of the Swiss administration have motivated building material producers and construction companies to invest in innovative EE housing technologies [84]. These advances in technology resulted in the formation of a voluntary, innovative EE standard—an implicit best practice in the architectural industry. This innovative standard had a technological advantage compared to the normal standard of constructing houses at that time [85]. The technical advantage improved the relative attractiveness of innovative EE housing. This activated the reinforcing loop R1: an increasing demand for innovative EE housing results in more innovative

EE housing—with the obvious qualification that the demand for innovative EE housing was small in absolute terms when compared to the demand for normal housing. Nevertheless, with new innovative EE housing becoming available, both producers of building material and construction companies and customers gained experience with this type of housing. The familiarity and the cost advantage of this innovative EE housing improved, and resulted in a higher demand. The higher demand for this type of housing motivates the generation of more EE construction capacity to fulfill this demand. Companies offer additional supply to meet the demand (B3). Both R1 and B3 mutually influence each other;

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

they both enable and constrain their development due to associated delays. For low values of innovative EE housing, the incumbent agents did not perceive the evolving innovative EE standard as a threat with the result that they restrain from protective actions. Hence, they did not initiate reductions in their production costs (B4) and they did not also reduce the existing support for innovative EE (R2a, R2b, R3). In other words, the respective loops were not active. Because of this, only the agents owning the capacity for constructing innovative EE housing supported the concept of innovative EE. Hence, the respective overall support was rather low. In addition, the willingness of political agents changed only with significant delays and seemed to be strongly nonlinear, i.e., changes were initiated only when extreme pressure existed, no change in the legal building code occurred at first. Steady incremental improvements in the relevant housing technologies continued to strengthen the technological advantage of the innovative standard. This occurred in small steps to enlarge the relative advantage of innovative EE housing, with its corresponding subsequent increase in demand as well as construction capacity. After the capacity for constructing innovative EE housing had improved considerably, and higher values of innovative EE housing had clearly demonstrated the technical and economic feasibility of innovative EE housing, the willingness of the political agent to improve the EE of the legal building code also increased. Incumbent agents soon realized the threat of innovative EE housing on their actual business: loops B4, R2, and R3 gained in strength. B4 reduced the relative cost advantage and also attractiveness of innovative EE housing and hence delayed and slowed R1; R2 and R3 reduced the support for innovative EE. Further research and developments of relevant housing technology continued to improve the technological advantage of innovative EE housing, and elaborated niches where normal EE housing technology could not compete. This technological advantage fueled R1 and, over time, overcame the counter-actions of loop B4. The corresponding construction capacity increased as well. The support for innovative EE combines the efforts of several agents. In the year 1982, this support crossed a threshold. The threshold depended on the strength of the support from innovative agents in the building construction industry (B5) relative to the strength of incumbent agents in the building sector (R2a), building construction industry (R2b), and the energy sector (R3). Willingness was then high enough to allow for improvements to the legal building code. The loop B1 became active and the EE of the legal code was increased with reference to the innovative standard. Improvements in the legal code reduced the technological advantage of the innovative EE standard. Due to both the improvements of the legal building code and the reduced technological advantage of the innovative standard, research institutions and manufactures were motivated to continue inventing and developing new innovative EE technologies. With delays for research and development, improved technologies became available on the market, which enabled higher levels of EE, which have been used to advance the level of EE of the innovative standard. These technological improvements also had a nonlinear effect on existing innovative EE housing and innovative

9

construction capacity. They reduced the lifetime of housing technology, leading to a degradation of existing innovative EE housing and construction technologies. In other words, the availability of new “state of the art” technologies reduced the previous innovative technologies to the normal category—the stocks of the first became smaller, whereas the stocks of the second became larger. With this, R1 lost its strength, since no innovative EE housing (according to the improved innovative EE standard) was available. Thus, both cost advantage and familiarity with the new technology returned to low levels; no demand existed for the new innovative EE housing—the dynamics of R1 are a helpful indicator for the status in an IDS-cycle. The end of an IDS-cycle seems to be indicated when the experience with innovation EE housing returns to low levels. To summarize, the interaction of seven feedback loops (Fig. 4, without B2) is used to explain the core dynamics of the first IDS-cycle in Switzerland. By now, in terms of the finite IDS-cycle, these core dynamics have played out fully once, leading to innovation, diffusion, and standardization of an EE building standard. To follow, I argue in less detail that the IDS-cycle has reoccurred multiple times from 1970 to 2010/11, and has thereby steadily improved the EE in the Swiss residential building codes. 4.2. Recurrence of the core dynamics of innovation–diffusion– standardization (IDS) In Switzerland, the improvement of the EE in building codes dates back to 1970. From a qualitative perspective, the IDS-cycle seems to have occurred five times, and has thus resulted in improvements of the EE in the legal building code in 1982, 1986, 1992, 2000, and 2008 [49,50,86–89].8 Throughout each single IDS-cycle, the loop B2 was not active and did not limit improvements of the legal building code. In other words, improvements to the legal code were contingent mostly by the respective willingness of political agents. However, as the EE of the legal building code improves, the marginal benefits from further improvements of the legal code's EE become smaller; thus, B2 becomes more significant and hence delays further improvements of the legal code. The fact that the time period between these improvements was augmented from 4 to 6 years and subsequently to 8 years (see Fig. 1) is empirical evidence for the effect of B2. Experts in the interviews have stated that this effect has started to impact the improvement process. One such alternative project to attract political agents would be to improve the EE of the Swiss energy generation and transmission system [90]. From the available data, it is not possible to ascertain whether loop B2 is already in full effect— although it will certainly be in the future. The political interest devoted to improving the EE of the legal building code will dwindle when EE has reached high levels. To summarize, the historical development of the EE in the Swiss residential sector can be explained by a recurring cycle of innovation, diffusion, and standardization. Each IDS-cycle

8 Without a quantitative simulation, this cannot be ascertained. As will be discussed in Section 6, this is a limitation of the qualitative system dynamics approach.

10

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

culminates in an improvement of the legal building code. From 1970 to 2010, arguably five such ISD-cycle iterations occurred, which can explain the historical co-evolution of EE in the voluntary innovative standard and the legal building code. With this, I have arrived at a structural model (Fig. 4) which is based on the historical development of the insightful case in Switzerland. To follow, I use that model to argue the possible effects of policies aimed at further advancement of EE in residential building codes. 5. Policy analysis The model introduced in the previous sections illustrated a situation with economic, social, technological, and political aspects. The following points are the analyses of policy interventions undertaken by the Swiss administration. The overarching objective of the Swiss administration is to improve the EE of the residential building stock and to reduce its energy demand and GHG emissions [82]. Administrative interventions are incorporated through feedback policies or changes of policy levers. First, I describe the additional information feedback and its likely impact on the development of EE, and then I address specific policy levers. Table 1 summarizes the feedback policies and levers. 5.1. Analysis of feedback policies A feedback policy is a decision rule that builds an endogenous cycle of causality. The introduction of a feedback policy uses information from a system state for decisionmaking and also tries to alter that state of the system. Fig. 5 shows the complete policy model. Based on the current EE of the building stock and its consequent annual energy demand, the administration forms its annual energy demand target. Here, I abstract from explicitly considering external factors to the residential building system, e.g., climate change reports [91,92], policy measures in other parts of the building

environment, e.g., public buildings, or developments in other countries, e.g., developments in the EU. These aspects are accounted for by external pressures on EE which influence the political target of annual energy demand. In this paper, I focus on causes endogenous to the Swiss residential building sector. The Swiss administration has followed the policy objective of reducing energy demand since the oil price shocks and the subsequent increasing public awareness in the 1970s. In 1998, the vision of a 2000 Watt Society which was also influenced by developments in the EU was formalized, specifically in countries such as Germany and the Netherlands, but also the USA, i.e., external pressures for EE. It requires a significant reduction in energy demand by 2050 [93–95]. When comparing the current level of EE with the target, the administration perceives a shortfall and concludes that, in order to meet that target, policy interventions are necessary [96]. As a consequence, the administration has begun to subsidize research projects to accelerate the rate of technological progress and to improve the EE of available housing technologies [9,96–100]. Technologies with higher levels of EE promote the creation of more innovative EE standards which, when used in construction and refurbishment, improve the average EE of the building stock. Thus, the intended effect of loop B6 is to reach the administration's objective. However, the introduction of this feedback policy into the existing building environment might also cause unintended consequences. One consequence may be that pushing technology development results in faster innovation cycles; the technology incorporated in the housing material and the assets of construction capacity lose their degree of novelty faster (policy 1, Table 1). For example, newly purchased innovative EE housing would be perceived as innovative for a period of 10–15 years if there was no policy intervention. With policy intervention, this period of novelty might be halved, because improved innovative EE technology becomes available faster. This would reduce the stock of innovative EE housing and thereby weaken the base of the growth engine for R1:

Table 1 Policy mechanisms (loops) and policy levers (parameter). #

Type of policy

Loop/variable

Description

Estimated feasibility

Impact on energy efficiency Low

1

Feedback policy

B6: Administration subsidizes R&D to improve energy efficiency of innovative standard

2

Feedback policy

3

Feedback policy

B8: Administration subsidizes already existing innovative energy efficient housing B7: Administration influences the political agenda

4

Parameter policy, lever Parameter policy, lever

Protective/innovative actions

Parameter policy, lever Parameter policy, lever

Cost advantage and familiarity of innovative EE housing Demolition rate of capacity for constructing normal EE housing and/or demolition rate of normal EE housing

5

6 7

Improving EE of building stock

Support R&D and improve the energy efficiency of innovative housing technologies; long delay until technology becomes available Support of innovative energy efficient housing currently available on the market; only short delay, direct effect Setting of low energy demand as target advances energy efficiency on the political agenda Support and compensate incumbent agents to reduce their resistance Support refurbishments of normal ee housing by financial means and information Charge carbon tax/levy on energy demand, increase energy price Selective destruction of construction capacity with low ee; selective destruction of normal energy efficient housing

High

High

Low

Medium Significant delay

No significant delay Significant delay

Low

Significant delay

Medium

Significant delay

Medium

Some delay

Low

High

Some delay

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

demand for innovative EE housing would be lower, less innovative capacity would be ordered, and the support for innovative EE would be lower with the result that the legal building code would be more improved as without the policy. Consequently, the EE of the building stock would not be improved, leading to a larger expected gap in annual energy demand. The administration would react to this with even stronger support, which would accelerate the dynamics (R4b); R4a follows the same dynamics, but for the capacity required for constructing innovative EE housing. A second feedback policy is to subsidize the purchase of innovative EE housing. This increases attractiveness and the demand for this type of housing (policy 2, Table 1). Consequently, more relevant capacity would be built, more support for innovative EE would be generated, and the legal building code would be improved exceeding what it otherwise would have been. The consequence would be that the EE of the building stock would increase with new construction and refurbishments; thus, the administration could come closer to its target (B8). However, this policy has a considerable potential to backfire. Incumbent agents can perceive the improvement of relative attractiveness in the innovative standard as a threat, and react to protect their current position in the industry, e.g., by increasing the economic efficiency of their production processes or incrementally improving their offerings. This would reduce the relative attractiveness of innovative EE housing. This could also mean that the support for innovative EE decreases and the development of the legal building code could freeze. Then incumbents might continue to improve their offerings to reduce the technical advantage of innovative EE housing to lower levels (R3). The administration would not accept this stagnation in building code development, since a delay or cessation of improvements in the building code would result in a gap in EE. The administration might react to a stagnating development with even stronger policy interventions to improve the economic attractiveness of innovative EE housing. As a consequence, the incumbent agents would themselves react even more strongly and try to counteract this push in policy (R5). The interactions of both R3 and R5 have the potential to unleash an escalation of action and reaction which could result in suboptimal economic efficiency. A further policy of the administration (policy 3, Table 1) might be that the political target of the annual energy demand reduces the marginal benefits of alternative investment projects, which would redirect the attention of political agents to the issue of EE in the legal building code (B7). By means of this policy, the constraining condition on loop B2 could be relaxed, thereby enabling further improvements in the legal building code. This would help the administration to come closer to its target. As the model shows, there are no unintended consequences with policy B7. However, this is largely because the model's boundary does not explicate other areas of the residential building sector. Future research might address this limitation. 5.2. Analysis of parameter policies With the complex network of feedback loops available (Fig. 5), I can now analyze the effect which changes in policy

11

levers might involve. What I understand as policy levers others have termed as pressure points [32]: they show options for interventions by the administration. In the following, I introduce four changes in policy levers and discuss their possible results. One policy lever is to address the protective actions of incumbent agents (policy 4, Table 1). This could be achieved by providing compensation for possible losses due to changes in the business landscape. When suitable compensation is found, the incumbent agents might not react as strongly to policy interventions by the administration; hence, the loop B4 would lose strength. With this, the escalating R3 and the R5 could also be reduced, enabling improvements of EE in building codes. However, the feasibility of this leverage depends strongly on the design of a suitable compensation scheme, which may be difficult. Aside from compensation for losses in normal business of incumbent agents, the scheme might also include stimulus packages motivating them to re-formulate their business models, e.g., in a way that utilities provide innovative energy services in the future. A second policy lever might be to improve the renovation rate of existing buildings [101]. This would improve the EE of the building stock. The stock of innovative EE housing would be increased with buildings from the stock of normal EE housing, given that renovations fulfill the innovative standard. This would weaken loop R2a and would help to improve the legal building code; further, it would strengthen R1. It might, however, strengthen B4, and the escalation of R3 and R5 would occur, leading to long delays and only minimal improvements in EE—a result not intended by the policy maker. Also, the feasibility of significantly increasing the renovation rate is only modest [102]. A third policy lever is to increase the energy price using a carbon tax or levy (policy 6, Table 1). This change would strengthen the loop R1 and in sequence also would activate B5: more demand for innovative EE housing would be generated, more capacity would be installed, and the legal code would be improved. Incumbent agents would, again, react with B4 and try to reduce the cost advantage of innovative EE housing. Also R3 and R5 could escalate, leading to ineffective use of resources and stagnating development of standards. Based on the diagram, I conclude that the impact on this policy change would be minimal in the short-term, since people would only regulate their energy consumption behavior. With a significant delay, however, customers would then opt for innovative EE housing. To harness these effects of the policy, it seems necessary to sustain the change in policy over a longer period. I assume that this is only moderately feasible. The last policy removes physical assets from the normal EE housing stock and also/or from the capacity for constructing normal EE housing (policy 7, Table 1). This deliberate reduction requires rebuilding housing along with housing construction capacity, which would comply with the innovative EE standard for both. This policy would weaken R2a and R2b; R1 would gain in strength as well as B5, which might result in an improvement of the EE in the legal building code. This would contribute to achieving the administration's target. In consequence, B6 and also B4 would become weaker. This policy would result in a considerable increase in the EE of the building stock with the fewest unintended consequences. However, I assume that the feasibility of this policy is rather low.

12

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

To conclude, this section has shown how the analysis of interacting feedback loops can help to uncover possible intended and unintended consequences. It seems that deliberately reducing the amount of normal EE housing or the corresponding construction capacity could significantly improve the EE in the building codes and the building stock, but the feasibility of this policy seems rather low. In the next section, I discuss insights about the specific case in Switzerland, the practical and theoretical implications, and finally the paper's limitations. 6. Discussion 6.1. Insights about the specific case of Switzerland I have addressed the phenomenon of how improvements of the EE codified in building standards are managed. The research has used the fact that the evolution of the EE in the building codes in Switzerland is different from other countries. The EE in Swiss residential building codes improved during the 1970s and early 1980s, but has remained the same since then, with little improvement during the 1990s, e.g., compared to the United States [41], United Kingdom [40], and Sweden [39]. The Swiss case has offered a successful example of the development of EE in building codes. Common explanations for the stagnation of improvements are caused by low energy prices and low economic pressure, or because the costs and benefits of innovative EE housing have been ineffectively distributed between actors in the residential sector [39,103]. In the Swiss case, the improvements of EE did not occur because of, but despite, declining pressures on energy prices (Fig. 1). From the analysis, I can conclude that by responding to the oil price shock and setting EE as a long-term political target [82] a cycle of innovation, diffusion, and standardization (IDS) has formed which has resulted in building codes with high EE. Particularly crucial was the formation of an innovative EE standard, which, from 1970 to 1980, developed as an implicit best practice in the industry. Because of this implicit standard, the legal building code could co-evolve with the implicit, voluntary standard. Evidence suggests that the IDS-cycle of improving the legal building code has recurred about five times up to today [85,87–89]. The difference to other countries, where improvements in EE of building standards have stalled when energy prices returned to low levels, is that in Switzerland the co-evolution of both standards have resulted in a productive pressure toward mutual development. 6.2. Implications for future policy interventions The policy analysis has shown that interventions to improve the level of the EE of building codes can also result in outcomes contrary to what was intended. The policy insights can be applied either to the Swiss case or to countries which have currently lower levels of EE in either building stock or building codes. For Switzerland, I have focused on the policy levers, since the feedback policies have already been established. The policy of increasing the energetic refurbishment of normal EE housing (policy 5, Table 1) is often regarded as successful in improving the EE of the building stock. However, the policy

might increase resistance from the incumbent agents who could try to reduce the attractiveness of innovative EE housing, e.g., reducing prices for normal technology, services and housing, thereby winning price-sensitive customers. These actions could actually result in a lower level of innovative EE housing than would be the case without policy intervention. In addition, their lobbying activities could lead to stagnation in the development of the standard. The resulting policy actions of the administration and the incumbents' counteraction could lead to wasteful escalating dynamics with a high risk to consume financial resources without adequate economic results. Actually, the resistance of the incumbent agents seems to be a central concern throughout most policy analyses, a factor which should be addressed adequately. One way of doing so would be to develop a compensation scheme for those agents. The design of this scheme would need to ensure an adequate rate of return for their current investments and a financially attractive stimulus to renew their assets. However, the scheme should not be too favorable for incumbents, thus leading to weaker innovative motivation. A policy with a large impact on the EE of the building stock and the level of EE in building codes would be to actively demolish normal EE housing capacity and the corresponding construction capacity, and to replace it with their innovative EE counterparts (policy 7, Table 1). This policy would lead to large and fast learning effects in the industry, so costs of innovative EE housing could drop significantly. Such a radical improvement would be difficult to counter by incumbent agents who would be more likely to concentrate on exploring new opportunities than competing on lost ground. The feasibility of this policy, however, seems low, since massive financial investments by the administration would be required. However, it might be possible to include the incumbent agents in this process of systemic transformation. Other countries can benefit from insights into the Swiss case. If the EE of a country's building stock is low, the administration might set a long-term reliable vision to signal to the other agents its commitment to improve EE and to subsidize research in technology (policy 1, Table 1). In addition to this, the administration might anticipate the resistance of incumbent agents, incentivizing as required (policy 4, Table 1) and including them in the development of a long-term solution. The paper's contribution is less a ready-to-use account to be used in other countries, but rather a specific language or building plan to construct dynamic models of other cases. In a more general sense, the use of causal policy models, which are able to indicate and communicate the dynamic consequences of policies, can uncover unintended consequences as well as provide means for discussing and solving problems to design policies for the future. 6.3. Theoretical implications The case analysis can complement existing theory. The literature on innovation diffusion as well as innovation and standardization can benefit from the endogenous perspective of the evolution of standards which the paper provides. Both diffusion and standardization research often assumes that the co-evolution of characteristics of standards and of technologies is exogenous [e.g., 10,55,104–106]. As the Swiss case

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

has shown, the level of improvement of housing technologies actually depends on the diffusion and standardization of previous housing technologies [similar for other industries, see 20]. The model offers a rich set of structural hypotheses which mathematical diffusion research can draw from. For example, the paper's model can provide a methodological perspective to recent research on standardization cycles [19]. For the literature on dominant design, this research demonstrates an insightful case where multiple designs, i.e., the legal building code and the voluntary innovative standard, exist simultaneously and do not compete with one another. In this respect, my work connects to work done by de Vries et al. [107]. Moreover, the case shows that both standards and designs are required to improve the overall level of performance of the industry [20]. In other words, multiple designs co-exist not only in competition, but also co-evolve in symbiosis. Furthermore, co-evolution research can draw on the results of the paper, which provides an additional example for a long-term co-evolution of institutions; in addition, it also enriches the addressed entities which co-evolve. In this paper, two legal institutions mutually influence each other, whereas other existing research has often concentrated on the co-evolution of a physical and a non-physical entity. And finally, the paper sits squarely within the literature on innovation systems research, which often has conceptualized institutional transition processes [36,60,108–110]. With the research, I provide a structural feedback model to explain a long-term dynamic phenomenon. With this, I provide a contribution to how innovation systems analysis could be operationalized by using an elaborated qualitative model. Thereby, I speak to similar efforts in innovation systems research [e.g., 11,26,60]. I also try to clarify how the innovation systems approach can serve as a basis for generating hypotheses, since the approach is still associated with conceptual diffuseness [111]. From a methodological perspective, the paper has broadened the existing qualitative system dynamics approach by explicitly formulating stock and flows. With this, I have increased the insights only possible from a qualitative analysis, at the cost of modest time investments compared to a quantitative analysis. Still, the certainty of the insights gained from a qualitative analysis are not comparable to the insights gained from a quantitative analysis; especially because in a qualitative analysis the behavior of a loop or a model can only be inferred mentally and therefore imprecise results can occur [28,29,112]. Readers should recall that I employ a qualitative analysis which is based on a formal conceptual model when evaluating the results. It is not feasible with the qualitative system dynamics approach to derive timing-based interventions or understand all possible counter-intuitive effects that a multiple interrelated feedback system can generate. For that purpose, a quantitative simulation is indispensible. 6.4. Limitations The work has several limitations. First, I address only the residential sector of the residential building sector and exclude the sectors that involve administrative, commercial, and office buildings. Field research indicates that the decisions and mechanisms of commercial and office building agents differ significantly from those of agents in the residential sector.

13

Second, I have limited the work to researching the evolution of the installed EE of residential buildings as defined by building standards, rather than a building's overall EE, which also depends largely on the behavior of the user [113–115]. It would not be possible to analyze the behavior of energy users as far back as 1970 for this selected case. Finally, the concept of relative advantage in technology does not account for co-benefits. Co-benefits are characteristics that accompany innovative EE housing, e.g., higher air quality, noise reduction, or low temperature heating. It is uncertain whether co-benefits are perceived by the customers as additional value or if customers consciously take them into account when deciding about the EE of a building [116,117]. 7. Concluding remarks A model was developed to explain the co-evolution of EE in building standards in the residential building sector. I used the revelatory case of Switzerland, which enabled us to develop a rich, qualitative model about innovation, diffusion, and standardization dynamics, and how their recurrence can explain the improvements of EE in the Swiss building codes from 1970 to 2010. The integrative model accounts for economic, social, technological, and legal factors, and shows that the co-evolution of both the legal building code and the voluntary innovative building standard create a regime which enables steady innovation even when the economic context would suggest otherwise. Future research can take several paths. One would be to operationalize the structural theory as a quantitative, feedback policy-analysis tool. This tool would enable one, first, to harden the counter-intuitive insights I have derived from the qualitative analysis, and second, to perform a policy analysis that also accounts for timing-based policy interventions. A second path would take the grounded model and expand it to other areas, for instance, with administrative and office buildings, or in other industries, e.g., in the ship-building, electricity, automobile, or information and telecommunication industry [66,118,119]. A third path would be to provide a systematic forecasting view on the standardization in the building environment and building codes. To this end, the methodology developed by Goluchowicz and Blind [120] could be used. A final path would be to generalize the model based on several case studies and to develop a more generic model for the evolution of standards. In all three cases, the structural model developed here can serve as a starting point. The quest to continuously improve the EE in standards and consequently in the building stock is a major challenge for policy makers [75,86,117,121,122]. Mastering this challenge can avoid causing economically harmful policy resistance. The qualitative model of the Swiss case was a first step toward developing a mathematical policy model which can computationally capture the dynamic complexity of the residential sector and help policy makers in decision-making. Leaving out the dynamic complexity of the system, as is often done, would result in oversimplified models with idealistic assumptions, such as perfect information or market efficiency. Such models might give rise to blind confidence, with the likely results of unrealistic forecasts, excessive investments in dead-end technologies, or accelerated commoditization of EE innovations [123]. To this end, an endogenous understanding of the

14

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

dynamic complexity of the residential building sector is required [31]—a necessity not many policy models can fulfill.

Acknowledgments I would like to thank Markus Schwaninger, Henry B. Weil, and Silvia Ulli-Beer as well as three anonymous reviewers of this journal, the reviewers of the International System Dynamics Conference 2011 in Washington, and the reviewers of the Academy of Management Conference 2011 in San Antonio for their helpful comments. This publication was supported by the Swiss National Science Foundation (PBSGP1_133613). In addition, I am grateful to the System Dynamics Group at the Massachusetts Institute of Technology for their support.

Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.techfore.2014.05.014.

References [1] IPCC, Summary for policymakers, in Climate Change 2007: the physical science basis, in: S. Solomon, D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor, H.L. Miller (Eds.), Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2007. [2] A.P. Sokolov, P.H. Stone, R. Prinn, M.C. Sarofim, M. Webster, S. Paltsev, C.A. Schlosser, C. Kicklighter, S. Dutkiewicz, J. Reilly, C. Wang, B. Felzer, H.D. Jacoby, Probabilistic forecast for 21st century climate based on uncertainties in emission (without policy) and climate parameters, Report Series of the MIT Joint Program on the Science and Policy of Global Change, Massachusetts Institute of Technology, Boston, 2009. [3] O. Edenhofer, R. Pichs-Madruga, Y. Sokona, K. Seyboth, D. Arvizu, T. Bruckner, J. Christensen, J.M. Devernay, A. Faaij, M. Fischedick, B. Goldstein, G. Hansen, J. Huckerby, A. Jäger-Waldau, S. Kadner, D. Kammen, V. Krey, A. Kumar, A. Lewis, O. Lucon, P. Matschoss, L. Maurice, C. Mitchell, W. Moomaw, J. Moreira, A. Nadai, L.J. Nilsson, J. Nyboer, A. Rahman, J. Sathaye, J. Sawin, R. Schaeffer, T. Schei, S. Schlömer, R. Sims, A. Verbruggen, C. von Stechow, K. Urama, R. Wiser, F. Yamba, T. Zwickel, Summary for policy makers, IPCC Special Report on Renewable Energy Sources and Climate Change Mitigation, Cambridge University Press, Cambridge, UK and New York, USA, 2011. [4] F. Meijer, L. Itard, M. Sunikka-Blank, Comparing European residential building stocks: performance, renovation, and policy opportunities, Build. Res. Inf. 37 (2009) 533–551. [5] H. Hens, G. Verbeeck, B. Verdonck, Impact of energy efficiency measures on the CO2-emissions in the residential sector: a largescale analysis, Energy and Buildings 33 (2001) 275–281. [6] M. Sunikka, Energy efficiency and low-carbon technologies in urban renewal, Build. Res. Inf. 34 (2006) 521–533. [7] M. Levine, D. Uerge-Vorsatz, K. Blok, L. Geng, D. Harvey, S. Lang, G. Levermore, A.M. Mehwana, S. Mirasgedis, A. Novikova, J. Rilling, H. Yoshino, Residential and commercial buildings: contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, in: B. Metz, O.R. Davidson, P.R. Bosch, R. Dave, L.A. Meyer (Eds.), Climate Change 2007: Mitigation, Cambridge University Press, Cambridge, UK and New York, USA, 2007. [8] McKinsey & Company, Impact of the Financial Crisis on Carbon Economics: Version 2.1 of the Global Greenhouse Gas Abatement Cost Curve, McKinsey & Company, 2010. [9] S.N. Groesser, Co-evolution of Standards in Innovation Systems: The Dynamics of Voluntary and Legal Building Codes, Springer, Heidelberg, 2012. [10] A. Higgins, G. Foliente, C. McNamara, Modelling intervention options to reduce GHG emissions in housing stock — A diffusion approach, Technol. Forecast. Soc. Chang. 78 (2011) 621–634.

[11] M. Dijk, M. Yarime, The emergence of hybrid-electric cars: innovation path creation through co-evolution of supply and demand, Technol. Forecast. Soc. Chang. 77 (2010) 1371–1390. [12] R. Guseo, M. Guidolin, Market potential dynamics in innovation diffusion: modelling the synergy between two driving forces, Technol. Forecast. Soc. Chang. 78 (2011) 13–24. [13] L.P. Bucklin, S. Sengupta, The co-diffusion of complementary innovations — supermarket scanners and UPC symbols, J. Prod. Innov. Manag. 10 (1993) 148–160. [14] M. Merito, A. Bonaccorsi, Co-evolution of physical and social technologies in clinical practice: the case of HIV treatments, Res. Policy 36 (2007) 1070–1087. [15] F. Murray, Innovation as co-evolution of scientific and technological networks: exploring tissue engineering, Res. Policy 31 (2002) 1389–1403. [16] D. Consoli, Co-evolution of capabilities and preferences in the adoption of new technologies, Tech. Anal. Strat. Manag. 20 (2008) 409–425. [17] K. Blind, S. Gauch, Research and standardisation in nanotechnology: evidence from Germany, J. Technol. Transfer. 34 (2009) 320–342. [18] B. Bozeman, Technology transfer and public policy: a review of research and theory, Res. Policy 29 (2000) 627–655. [19] S. Botzem, L. Dobusch, Standardization cycles: a process perspective on the formation and diffusion of transnational standards, Organ. Stud. 33 (2012) 737–762. [20] K. Blind, Driving forces for standardization at standardization development organizations, Appl. Econ. 34 (2002) 1985–1998. [21] K. Blind, Explanatory factors for participation in formal standardisation processes: empirical evidence at firm level, Econ. Innov. N. Technol. 15 (2006) 157–170. [22] B.A. Lundvall, B. Johnson, E.S. Andersen, B. Dalum, National systems of production, innovation and competence building, Res. Policy 31 (2002) 213–231. [23] M.P. Hekkert, S.O. Negro, Functions of innovation systems as a framework to understand sustainable technological change: empirical evidence for earlier claims, Technol. Forecast. Soc. Chang. 76 (2009) 584–594. [24] M.P. Hekkert, R.A.A. Suurs, S.O. Negro, S. Kuhlmann, R.E.H.M. Smits, Functions of innovation systems: a new approach for analysing technological change, Technol. Forecast. Soc. Chang. 74 (2007) 413–432. [25] F. Kern, Using the multi-level perspective on socio-technical transitions to assess innovation policy, Technol. Forecast. Soc. Chang. 79 (2012) 298–310. [26] S.O. Negro, M.P. Hekkert, Explaining the success of emerging technologies by innovation system functioning: the case of biomass digestion in Germany, Tech. Anal. Strat. Manag. 20 (2008) 465–482. [27] J. de Haan, J. Rotmans, Patterns in transitions: understanding complex chains of change, Technol. Forecast. Soc. Chang. 78 (2011) 90–102. [28] R.G. Coyle, Qualitative and quantitative modelling in system dynamics: some research questions, Syst. Dyn. Rev. 16 (2000) (225-144). [29] M. Schaffernicht, Causal loop diagrams between structure and behaviour: a critical analysis of the relationship between polarity, behaviour and events, Syst. Res. Behav. Sci. 27 (2010) 653–666. [30] M. Schwaninger, S.N. Groesser, Model-based theory-building with system dynamics, Syst. Res. Behav. Sci. 25 (2008) 447–465. [31] J.D. Sterman, Learning in and about complex-systems, Syst. Dyn. Rev. 10 (1994) 291–330. [32] R.G. Coyle, System Dynamics Modelling: A Practical Approach, Chapman and Hall, London, 1996. [33] J.D. Sterman, Business Dynamics: Systems Thinking and Modeling for a Complex World, McGraw-Hill, Boston, MA, 2000. [34] R.K. Yin, Case Study Research, Sage Publications, Beverly Hills, CA, 2008. [35] K.M. Eisenhardt, Building theories from case-study research, Acad. Manag. Rev. 14 (1989) 532–550. [36] F.W. Geels, Processes and patterns in transitions and system innovations: refining the co-evolutionary multi-level perspective, Technol. Forecast. Soc. Chang. 72 (2005) 681–696. [37] A.L. Strauss, J.M. Corbin, Basics of Qualitative Research Techniques and Procedures for Developing Grounded Theory, 2nd ed., Sage Publications, Thousand Oaks, California, 1998. [38] AWEL, Energieplanungsbericht 2006: Bericht des Regierungsrates über die Energieplanung des Kantons Zürich, http://www.energie.zh.ch/ internet/bd/awel/energie/de/home.SubContainerList.SubContainer2. ContentContainerList.0007.DownloadFile.pdf?CFC_cK=1253261142962, Zurich 2007. [39] J. Nassen, J. Holmberg, Energy efficiency — a forgotten goal in the Swedish building sector? Energy Policy 33 (2005) 1037–1051. [40] S. Roberts, Altering existing buildings in the UK, Energy Policy 36 (2008) 4482–4486. [41] DOE, Residential Energy Consumption, access date: 23rd September 2009, http://www.eia.doe.gov/emeu/consumption/index.html 2009.

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16 [42] BP, Oil Price Statistics from 1871, access date: 23rd September 2009, http://www.bp.com/sectiongenericarticle.do?categoryId=9023773& contentId=7044469 2009. [43] M. Müller, S.N. Groesser, S. Ulli-Beer, How do we know who to include in transdisciplinary research? Toward a method for the identification of experts, Eur. J. Oper. Res. 216 (2012) 495–502. [44] E.M. Rogers, The Diffusion of Innovations, Free Press, New York, NY, 2013. [45] P. Stoneman, The Economics of Technical Diffusion, Blackwell Publishers, Oxford, 2002. [46] P. Stoneman, P. Diederen, Technology diffusion and public policy, Econ. J. 104 (1994) 918–930. [47] G. Dosi, Technological paradigms and technological trajectories — a suggested interpretation of the determinants and directions of technical change, Res. Policy 11 (1982) 147–162. [48] J.M. Utterback, Mastering the Dynamics of Innovation: How Companies Can Seize Opportunities in the Face of Technological Change, Harvard Business School Press, Boston, MA, 1994. [49] M. Jakob, Grundlagen zur Wirkungsabschätzung der Energiepolitik der Kantone im Gebäudebereich, Department of Energy, Zurich, 2008. [50] SFOE, Vorstudie zur Erhebung von Energiekennzahlen von Wohnbauten, Department of Energy, Bern, 2007. [51] M. Koschenz, A. Pfeiffer, Potenzial Wohngebäude: Energie- und Gebäudetechnik für die 2000-Watt-Gesellschaft, Faktor-Verlag, Zürich, 2005. [52] T.F. Schulz, Intermediate Steps Towards the 2000-Watt Society in Switzerland: An Energy-economic Scenario Analysis, Dissertation Swiss Federal Institute of Technology Zurich, 2007. [53] P. Biermayr, B. Baumann, E. Schriefl, H. Skopetz, K. Stieldorf, H. Zelenka, M. Berger, M. Ornetzeder, Analyse fördernder und hemmender Faktoren bei der Markteinführung von innovativen Wohnbauten: Endbericht, Institut für Energiewirtschaft, Wien, 2001. [54] P. Biermayr, E. Schriefl, B. Baumann, A. Sturm, Hemmnisse und fördernde Faktoren bei der Markteinführung innovativer Wohnbauten: eine Informationsoffensive für Planer, Wohnbauträger und Technologieproduzenten, “Innovationsoffensive”, Bundesministerium für Verkehr, Innovation und Technologie, Wien, 2002. [55] T. Siller, M. Kost, D. Imboden, Long-term energy savings and greenhouse gas emission reductions in the Swiss residential sector, Energy Policy 35 (2007) 529–539. [56] J.A.M. Vennix, Group Model Building: Facilitating Team Learning Using System Dynamics, Wiley, Chichester, 1996. [57] D.F. Andersen, G.P. Richardson, Scripts for group model building, Syst. Dyn. Rev. 13 (1997) 107–129. [58] S.N. Groesser, M. Schaffernicht, Mental models of dynamic systems: taking stock and looking ahead, Syst. Dyn. Rev. 28 (2012) 46–68. [59] J.W. Forrester, Industrial Dynamics, Productivity Press, Cambridge, MA, 1961. [60] F.W. Geels, M.P. Hekkert, S. Jacobsson, The dynamics of sustainable innovation journeys, Tech. Anal. Strat. Manag. 20 (2008) 521–536. [61] F.Y. Phillips, Market-oriented Technology Management: Innovating for Profit in Entrepreneurial Times, Springer, Heidelberg, 2001. [62] D. Urge-Vorsatz, A. Novikova, Potentials and costs of carbon dioxide mitigation in the world's buildings, Energy Policy 36 (2008) 642–661. [63] A. Audenaert, S.H. De Cleyn, B. Vankerckhove, Economic analysis of passive houses and low-energy houses compared with standard houses, Energy Policy 36 (2008) 47–55. [64] J. Ravetz, State of the stock—what do we know about existing buildings and their future prospects? Energ Policy 36 (2008) 4462–4470. [65] M.J. Koetse, A.J. van der Vlist, H.L.F. de Groot, The impact of perceived expectations and uncertainty on firm investment, Small Bus. Econ. Group 26 (2006) 365–376. [66] B. Dattée, H.B. Weil, Dynamics of social factors in technological substitutions, Technol. Forecast. Soc. Chang. 74 (2007) 579–607. [67] V. Venkatesh, M.G. Morris, G.B. Davis, F.D. Davis, User acceptance of information technology: toward a unified view, MIS Q. 27 (2003) 425–478. [68] K. Arkesteijn, L. Oerlemans, The early adoption of green power by Dutch households — an empirical exploration of factors influencing the early adoption of green electricity for domestic purposes, Energy Policy 33 (2005) 183–196. [69] R.H. Acker, D.M. Kammen, The quiet (energy) revolution — analysing the dissemination of photovoltaic power systems in Kenya, Energ Policy 24 (1996) 81–111. [70] J.C. Sweeney, G.N. Soutar, T. Mazzarol, Factors influencing word of mouth effectiveness: receiver perspectives, Eur. J. Mark 42 (2008) 344–364. [71] E. Abrahamson, L. Rosenkopf, Social network effects on the extent of innovation diffusion: a computer simulation, Organ. Sci. 8 (1997) 289–309.

15

[72] S. Kapur, Technological diffusion with social-learning, J. Ind. Econ. 43 (1995) 173–195. [73] K. Chatterjee, S.H. Xu, Technology diffusion by learning from neighbours, Adv. Appl. Probab. 36 (2004) 355–376. [74] H.P. Young, The spread of innovations through social learning, The Brookings Institution, CSED Working Paper No. 43, Johns Hopkins University, Washington, DC, 2005. [75] A.B. Jaffe, R.N. Stavins, The energy paradox and the diffusion of conservation technology, Resour. Energy Econ. 16 (1994) 91–122. [76] S. Alberth, Forecasting technology costs via the experience curve — myth or magic? Technol. Forecast. Soc. Chang. 75 (2008) 952–983. [77] A. McDonald, L. Schrattenholzer, Learning rates for energy technologies, Energy Policy 29 (2001) 255–261. [78] R. Henderson, Of life-cycles real and imaginary — the unexpectedly long old-age of optical lithography, Res. Policy 24 (1995) 631–643. [79] R.G. Coyle, A systems approach to the management of a hospital for short-term patients, Socio Econ. Plan. Sci. 18 (1984) 219–226. [80] FSO, Swiss Environmental Statistics: A Brief Guide 2008, Federal Office for the Environment, Berne, 2008. [81] OECD/IEA, IAE Energy Technology R&D Statistics 1975–1995, OECD/ IEA, Paris Cedex, 1997. [82] C. Azar, H. Dowlatabadi, A review of technical change in assessment of climate policy, Annu. Rev. Energy Environ. 24 (1999) 513–544. [83] D.H. Meadows, D. Meadows, J. Randers, W.W. Behrens III, The Limits to Growth, Universe Books, New York, NY, 1972. [84] D.M. Gann, Y.S. Wang, R. Hawkins, Do regulations encourage innovation? The case of energy efficiency in housing, Build. Res. Inf. 26 (1998) 280–296. [85] SFOE, Internationaler Vergleich von Energiestandards im Baubereich, Swiss Federal Office of Energy, Bern, 2005. [86] R. Meier, W. Ott, Grundlagen für eine Strategie Gebäudepark Schweiz, , Bundesamt für Energie, Bern, 2005. [87] K. Gantenbein, Building for the 2000-Watt Society: The State of Affairs, Stadt Zurich Hochbaudepartment, Zurich, 2009. [88] J.D. Delley, L. Mader, L'état face auf defi energetique, études et pratique, Payot, Lausanne, 1986. [89] EnDK, Mustervorschriften der Kantone im Energiebereich (MuKEn), http://www.endk.ch/kantone.html 2011 (accessed: 10th May 2011, Zurich). [90] Company & McKinsey, Swiss Greenhouse Gas Cost Abatement Curve, McKinsey & Company, 2009. [91] IPCC, Climate Change 2001: synthesis report — summary for policymakers, Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2001. [92] IPCC, Summary for Policymakers of the Synthesis Report of the IPCC Fourth Assessment Report, Intergovernmental Panel on Climate Change, Wembley, 2007. [93] E.T.H., Council, Strategie Nachhaltigkeit im ETH-Bereich: 2000 WattGesellschaft — Modell Schweiz, Wirtschaftsplattform, ETH Zurich, Zurich, 1998. [94] A. Pfeiffer, M. Koschenz, Energy and building technology for the 2000 Watt Society: potential of residential buildings in Switzerland, Energy Build. 37 (2005) 1158–1174. [95] F. Marechal, D. Favrat, E. Jochem, Energy in the perspective of the sustainable development: the 2000 W Society challenge, Resour. Conserv. Recycl. 44 (2005) 245–262. [96] M. Beerepoot, N. Beerepoot, Government regulation as an impetus for innovation: evidence from energy performance regulation in the Dutch residential building sector, Energy Policy 35 (2007) 4812–4825. [97] BFE, Nachbefragung und Erfolgskontrolle Investitionsprogramm Energie 2000: Schlussbericht, Swiss Federal Office of Energy, Berne, 2005. [98] BFE, EnergieSchweiz, Das Nachfolgeprogramm von Energie 2000, Swiss Federal Office of Energy, Berne, 2001. [99] BFE, Rationelle Energienutzung in Gebäuden: Aktivitäten und Projekte 2002, Swiss Federal Office of Energy, Berne, 2002. [100] M. Kaufmann, H.-P. Nützi, EnergieSchweiz in der 2. Etappe — mehr Wirkung, mehr Nutzen, EnergieSchweiz, Bern, 2005. [101] CCEM, CCEM Retrofit, Advanced Low Energy Renovation of Buildings, 2011. [102] R.W. Amstalden, M. Kost, C. Nathani, D.M. Imboden, Economic potential of energy-efficient retrofitting in the Swiss residential building sector: the effects of policy instruments and energy price expectations, Energy Policy 35 (2007) 1819–1829. [103] J. Nassen, F. Sprei, J. Holmberg, Stagnating energy efficiency in the Swedish building sector-economic and organisational explanations, Energ Policy 36 (2008) 3814–3822. [104] T.F. Schulz, S. Kypreos, L. Barreto, A. Wokaun, Intermediate steps towards the 2000 W Society in Switzerland: an energy-economic scenario analysis, Energy Policy 36 (2008) 1303–1317.

16

S.N. Groesser / Technological Forecasting & Social Change 87 (2014) 1–16

[105] R. Madlener, Innovation diffusion, public policy, and local initiative: the case of wood-fuelled district heating systems in Austria, Energy Policy 35 (2007) 1992–2008. [106] M. Weiss, L. Dittmar, M. Junginger, M.K. Patel, K. Blok, Market diffusion, technological learning, and cost–benefit dynamics of condensing gas boilers in the Netherlands, Energy Policy 37 (2009) 2962–2976. [107] H.J. de Vries, J.P.M. de Ruijter, N. Argam, Dominant design or multiple designs: the flash memory card case, Tech. Anal. Strat. Manag. 23 (2011) 249–262. [108] F.W. Geels, The dynamics of transitions in socio-technical systems: a multi-level analysis of the transition pathway from horse-drawn carriages to automobiles (1860–1930), Tech. Anal. Strat. Manag. 17 (2005) 445–476. [109] C. Marxt, C. Brunner, Analyzing and improving the national innovation system of highly developed countries — The case of Switzerland, Technol. Forecast. Soc. Chang. 80 (6) (2013) 1035–1049. [110] T. Daim, D. Kocaoglu, T. Anderson, Emerging frameworks describing technological innovation: review of multiple perspectives, Technol. Forecast. Soc. Chang. (2014) (in press). [111] C. Edquist, The Systemic Nature of Innovation, Oxford Handbooks Online, Oxford, 2006. [112] D.C. Lane, The emergence and use of diagramming in system dynamics: a critical account, Syst. Res. Behav. Sci. 25 (2008) 3–23. [113] P.C. Stern, Understanding individual's environmentally significant behavior, ELR News Anal. 35 (2005) 10785–10790. [114] S. Sorrell, J. Dimitropoulos, M. Sommerville, Empirical estimates of the direct rebound effect: a review, Energy Policy 37 (2009) 1356–1371. [115] P. Gyberg, J. Palm, Influencing households' energy behaviour: how is this done and on what premises? Energy Policy 37 (2009) 2807–2813. [116] J. Ansar, R. Sparks, The experience curve, option value, and the energy paradox, Energy Policy 37 (2009) 1012–1020.

[117] L. Weber, Some reflections on barriers to the efficient use of energy, Energ Policy 25 (1997) 833–835. [118] V.B. Kreng, B.J. Wang, An innovation diffusion of successive generations by system dynamics — an empirical study of Nike Golf Company, Technol. Forecast. Soc. Chang. 80 (2013) 77–87. [119] D.H. Lee, S.Y. Park, J.W. Kim, S.K. Lee, Analysis on the feedback effect for the diffusion of innovative technologies focusing on the green car, Technol. Forecast. Soc. Chang. 80 (2013) 498–509. [120] K. Goluchowicz, K. Blind, Identification of future fields of standardisation: an explorative application of the Delphi methodology, Technol. Forecast. Soc. Chang. 78 (2011) 1526–1541. [121] A.B. Jaffe, R.N. Stavins, The energy-efficiency gap — what does it mean, Energ Policy 22 (1994) 804–810. [122] W. Kägi, M. Schäfli, S. Siegrist, W. Hässig, Best Practice: Marktordnung, Markttransparenz und Marktregelung zugunsten der Durchsetzung energieeffizienter Lösungen am Markt, Department of Energy, Berne, 2004. [123] H.B. Weil, Why markets make mistakes, Kybernetes 39 (2010) 1429–1451. Stefan N. Groesser is Professor of Strategic Management at the Bern University of Applied Sciences, Switzerland. In addition, he is a Senior Lecturer for System Dynamics at the University of St Gallen, Switzerland, and a visiting scholar at the System Dynamics Group at MIT Sloan. He has received degrees from the University of Stuttgart, Germany, in business administration and economics, from the University of Bergen, Norway, in System Dynamics, and from the University of St. Gallen, Switzerland, in Management. Stefan's research interests include strategic management, business models, mental models, and simulation methodology.