Increasing the flexibility of electricity consumption in private households: Does gender matter?

Energy Policy 118 (2018) 9–18

Contents lists available at ScienceDirect

Energy Policy journal homepage: www.elsevier.com/locate/enpol

Increasing the flexibility of electricity consumption in private households: Does gender matter?

T

⁎

Lise Tjørringa, , Carsten Lynge Jensenb, Lars Gårn Hansenb, Laura Mørch Andersenb a b

Elholm 1, DK-6400 Sonderborg and Department of Food and Resource Economics, Faculty of Science, University of Copenhagen, Denmark Department of Food and Resource Economics, Faculty of Science, University of Copenhagen, Denmark

A R T I C LE I N FO

A B S T R A C T

Keywords: Gendered practices Flexible electricity consumption Households Energy Household appliances Field study

Increased reliance on renewable electricity production is important for reducing greenhouse gas emissions. Inducing households to adjust the timing of their electricity consumption to fit fluctuations in renewable electricity supply can help make this possible. In a field study, 71 Danish households received incentivized text messages, suggesting that they shift electricity consumption to certain hours of the day. The study shows that when text messages were sent to women, there was a significantly greater response than when they were sent to men. Based on qualitative in-depth interviews, we find that an important reason for this is gender difference in household work task responsibility. Our study suggests that incorporating knowledge about gendered practices when designing and targeting policies to change electricity consumption habits and induce flexibility could significantly increase their effectiveness.

1. Introduction Expanding renewable electricity production is an important part of the greening strategy of many countries (Committee on Climate Change, 2015; The Danish Government, 2011; The Government of Norway, 2012; The Government of Sweden, 2008). However, the availability of renewable wind and solar energy can change dramatically during the day, resulting in increased supply volatility (Chu and Majumdar, 2012; Karki and Billinton, 2001). In other words, the supply of renewable electricity is determined 'as the wind blows'. One way of addressing this is to induce greater flexibility in electricity consumption. According to the Danish Energy Commission, the Danish Energy Association and the state-owned company that owns the Danish electricity and gas transmission system, flexible electricity consumption will become an important element of a well-functioning future intelligent energy system (Altinget, 2016; Danish Energy Association and Energinet.dk, 2013). This also seems to be the expectation of the EU (European Commision, 2011). In the EU, private households account for 30% of total electricity consumption (European Environment Agency, 2015). Similarly, in Denmark, private households account for 31% of total energy consumption (Danish Energy Agency, 2016). Hence, developing flexible electricity consumption within private households has great potential. To date, promoting flexible electricity consumption within private

⁎

1

Corresponding author. E-mail address: [email protected] (L. Tjørring). In section 2, we explain how we perceive the concept of gender in this paper.

https://doi.org/10.1016/j.enpol.2018.03.006 Received 26 April 2017; Received in revised form 27 February 2018; Accepted 3 March 2018 0301-4215/ © 2018 Elsevier Ltd. All rights reserved.

households has mainly been through technological fixes and economic incentives, but these approaches have been criticized for downplaying the cultural dimensions of energy use (D’Agostino et al., 2011; Kempton and Layne, 1994; Lutzenhiser, 1992; Lutzenhiser and Shove, 1999; Sovacool et al., 2012; Wilhite et al., 2000). In order to fully understand flexible electricity consumption, we need to understand, on a very practical level, what different people do within private households and how this generates electricity demand (Anderson, 2016; Aune, 2007; Skjølsvold et al., 2016; Walker, 2014; Widén et al., 2009). Much of what people do within households is culturally determined according to gender (Munro and Madigan, 2006).1 In fact, the household is one of the most gendered spheres of society in most cultures. Despite many years of gender equality work, there is a significant gender division within the modern family home in western countries (ibid.). Household chores, such as cooking and doing the laundry, are predominantly performed by women (Carlsson-Kanyama and Lindén, 2007; Ellegård and Palm, 2011; Laermans and Meulders, 2006; Shove, 2003), and women usually spend much more time performing household chores than men (Bianchi et al., 2000; Carlsson-Kanyama and Lindén, 2007; Sayer, 2010). Because flexible electricity consumption requires changing the timing of daily electricity-consuming household chores to different times of the day, gender is likely to have an important mediating effect on how households respond to flexible electricity consumption programs.

Energy Policy 118 (2018) 9–18

L. Tjørring et al.

have different preferences and practices, and they have different positions in the household hierarchy (Ellsworth-Krebs et al., 2015). This is not to say that single households are not affected by cultural gender norms, or that they are not relevant to examine in the context of increasing the flexibility of energy consumption. It is merely an expression of our choice of an in-depth focus on gender and the interaction between people within the household. Focusing on households that consist of at least two people is not meant to be representative of all households. Rather, it is an explorative study to investigate the importance of gender when addressing energy consumption in private households. As West and Zimmerman (1987) point out, gender roles are created through human interactions. In this paper, we have chosen to focus on the connection between gender and activities in the house in a very concrete manner because energy consumption is closely connected to what people do in the house. People's daily practices essentially demand electricity in private households (Shove, 2003; Strengers, 2012). Using electricity in private households (and elsewhere) is only a medium for performing tasks such as the laundry, cooking, dishwashing, etc. When we turn on a lamp, it is not because we want to use electricity, but rather because we want light and we want to do something with the light. Therefore, in order to understand electricity consumption, we need to understand the practices surrounding using light, doing the laundry, cooking, etc. In this paper, we focus on how gender is connected to these energy-consuming household practices. We use the term, practice, which is understood as people's daily household activities such as cooking, doing the laundry, drying clothes, etc. The concept of practice is used as a descriptive term, rather than an analytical term with theoretical and abstract implications (see Pink, 2012:16).

Several studies have found that gender-specific practices are important for electricity consumption. Ellegård and Palm (2015, 2011) find that men and women carry out different activities in the home and, hence, consume energy in different ways. Clancy and Roehr (2003) point out that women often take on the responsibility for reducing electricity use connected with household electric appliances. In line with this, Carlsson-Kanyama and Lindén (2007) find that it is mainly women who respond to increases in peak hour electricity tariffs because they perform tasks such as washing clothes and dishes, which can be shifted to low peak hours. Women also avoid using tumble dryers, which results in more time being spent on this task. Jensen et al. (2012) find that single women are less likely to install auto-power-off devices, which reduce standby electricity consumption, than single men and couples. They suggest this might be due to gender differences in the skills and/or interest in installing the technology and willingness to spend time and effort on installation. Tjørring (2016) shows that the different practices of men and women in the house influence the negotiation and decision-making process for doing energy renovations. Because men and women do different things in the shared household, they prefer to invest in different forms of renovation that affect their particular practices (ibid.). Generally, women influence and are influenced by energy issues in different ways to men, whether it be due to differences in the way men's and women's everyday lives are influenced by energy conservation (Carlsson-Kanyama and Lindén, 2007), unequal representation of women and men in the energy sector (Carlsson-Kanyama et al., 2010; Clancy and Roehr, 2003; Ryan, 2014), differences in the use of energy (Ellegård and Palm, 2015, 2011; Isaksson and Ellegård, 2015; Räty and Carlsson-Kanyama, 2010; Tjørring, 2016), or differences in environmental concern (Clancy and Roehr, 2003; Dietz et al., 2002; Zelezny et al., 2000). Knowledge about these gender differences can be used to develop targeted, more energy-efficient energy supply and demand. In this paper, we investigate the effect gender has on private households' participation in a Danish experimental flexible electricity consumption program. More precisely, through two questionnaires and twenty in-depth interviews with participants in the experimental flexible electricity consumption program, we investigate how the potentially gendered everyday energy consuming activities within the household affect the adoption of flexible electricity consumption. In the following, we start by explaining the social cultural model for understanding gender-specific differences in the work tasks related to the flexible electricity consumption of households. This is followed by an outline of the field study of a flexible electricity consumption program and the research method, which is based on a combination of questionnaires and qualitative in-depth interviews with the private households participating in the field study. Next, we present the results of the questionnaires and qualitative interviews and, finally, discuss the implications of gender differences for practice and policy.

3. The field study In a field study, the potential for increasing flexible electricity consumption through monetary and pro-social incentives was investigated.3 Fig. 1 shows the average daily electricity consumption in 2011 of the 247,010 private households in the study area supplied by the power company SE. The intervention involved giving households incentives to sometimes shift their power consumption to a time during the period 20:00 and 23:00 and sometimes to a time outside these hours (period between dotted lines). The power use profile indicates that many households use a lot of power-intensive appliances just prior to the period, but that use decreases during the period 20:00–23:00. Hence, there is potential to even out the peak by shifting electricity consumption to the period between 20:00 and 23:00. However, the households were also sometimes encouraged to shift their electricity consumption away from these hours. The reason for this was to simulate a situation in which the electricity consumption is adjusted to the production of wind energy. In this situation, electricity production alternates between there being too much and too little. The intervention period was static, i.e. it was always between 20:00 and 23:00. Another Danish research project on flexible electricity consumption had found that more households adopted flexible electricity consumption with static-time of use prices rather than real-time-prices (Friis and Haunstrup Christensen, 2016). Finally, the period between 20:00 and 23:00 was chosen based on the assumption that most households are at home during this period and, hence, have the opportunity to change their electricity consumption. A sample of 4042 households was randomly selected from SE's customer data base4 and invited to join what was called the MovePower

2. Theoretical perspectives: Social and cultural perspective on gender In this paper, the roles of men and women are seen as being socially constructed rather than biologically determined. The concept of gender is in line with West and Zimmerman's (1987) social constructivist perspective (West and Zimmerman, 1987). Furthermore, in this paper, gender is investigated in the context of a family consisting of a couple with or without children. These categories are based on the power company SE's use of Statistics Denmark's definitions.2 This opens up for recognizing the interaction between people who live together, which is likely to be affected by gender. As Ellsworth-Krebs et al. (2015) point out, there is a tendency to overlook the interaction between householders in the context of energy consumption. Household members 2

3 The field study was organized by a partnership between the energy company, SE, the University of Copenhagen, the Danish Technical University and the private company, Develco Products. 4 SE has a database of their 247,010 customers. Of the 247,010 customers, 40,490 have given SE contact permission. 4042 households were randomly selected and invited out of

There is some uncertainty as to how SE has implemented the definitions.

10

Energy Policy 118 (2018) 9–18

L. Tjørring et al.

Fig. 1. Hourly kWh use per household in SE's customer area, 2011 The dotted lines indicate the intervention hours between 20:00 and 23:00.

program. In April 2014, all selected households received an e-mail with an invitation, which asked whether they would be willing to be part of the new flexible electricity scheme. In this e-mail, all households were informed that they would receive text messages suggesting time slots for changing the timing of their electricity use and other general information (See Appendix). In order to sign up, the households had to register the gender, age and cell phone number of the person who they had chosen to receive the text messages. At the enrollment, the households were also presented with incentives to move the timing of their electricity use. Incentives were randomly varied across households with some receiving a rebate (of 0.50 kr., 1.00 kr. or 1.50 kr. per kWh electricity moved in accordance with the text messages), while others were given an environmental incentive, where SE promised to increase the generation of wind power (by 0.50 kWh, 1.00 kWh or 1.50 kWh per kWh electricity moved in accordance with the text messages). Finally, some households received both a rebate and an environmental incentive. The households did not receive any further monetary incentives to take part in the survey. This paper is based on a sub-sample of the field study investigating the effect gender had on private households' participation in the flexible electricity consumption program. The sub-sample consisted of 2780 households of couples or families with children. Single households were excluded from the gender sub-project as we were interested in the gendered interaction between household members. In total, 71 households from this sub-sample signed up to participate in the MovePower program. On an everyday basis, the new electricity scheme entailed receiving a text message containing information about what time of the day they should or should not use electricity and giving an incentive (monetary, environmental or both). Text messages were received the same day households were requested to change the timing of their electricity consumption. The time frame for changing the timing of the electricity consumption was always between 20:00 and 23.00. Thus, they were informed by text messages to either move their electricity consumption to/away from the period 20:00–23:00. Households were only supposed to change the timing of their electricity consumption on the days they received a text message alert. The first text messages were sent in May 2014, and the last text message was sent in June 2015.5 Information about the households' energy consumption was provided by the energy company. A summary of the statistics regarding

Table 1 Summary statistics regarding gender, age and kWh electricity consumption in the flexible electricity consumption program. Gender of the person in the household receiving the text messages

Number (N) of persons Mean (M) age Sample Standard Deviation (SD) of age Mean (M) annual kWh electricity consumption in households Sample Standard Deviation (SD) of annual kWh electricity consumption

Women

Men

33 58.16 2.19 3548

38 58.11 1.94 4161

282

261

gender, age and kWh electricity consumption of those who signed up to receive the text messages is outlined in Table 1. Two sided t-tests6 were conducted, which showed that there was no significant difference in age between the participating men and women and no significant difference in annual average electricity consumption between households where the woman had received the text messages and households where the man had received the text messages. The measurement of the electricity consumption was important to the extent that it indicated how many electrical appliances the households had and indicated the overall potential for changing the timing of the consumption. 4. Methodology: Two questionnaires and ten qualitative in-depth interviews We used a combination of questionnaires and qualitative in-depth interviews to investigate the effect of gender on the ability to increase flexibility in electricity consumption in private households. Table 2 below presents an overview of the methods. The participating households were invited to answer two online questionnaires; one at the start of the field study in April 2014, and the second at the end of the field study in June 2015. The questionnaires were sent out by email. The aim of the first questionnaire was to 6 A t-test is a statistical hypothesis test, which can be used to determine whether the mean value of two sets of data are significantly different from each other. The ages of women (M = 58.16, SD = 2.19) and men (M = 58.11, SD = 1.94) are not significantly different at the 5% level (t(67) = 0.02, p = 0.98, two-sided t-test). The annual kWh consumption of households where women receive the text messages (M = 3548, SD = 282) and of households where men receive the text messages (M = 4161, SD = 261) are not significantly different at the 5% level (t(69) = 1.59, p = 0.12, two-sided t-test).

(footnote continued) the group of 40,490. 5 The MovePower experiment was repeated on a larger scale the following year from June 2015 to June 2016 with 40,487 invited households and 2732 participants. Subsequently, the MovePower program ended.

11

Energy Policy 118 (2018) 9–18

L. Tjørring et al.

increased the flexibility of their electricity consumption and changed the timing of their practices of washing, dishwashing, tumble drying and using the cell phone, PC, etc. The practices of using air-conditioning and heating were ignored because air-conditioning is rarely used in Denmark and the source of heating is mainly gas, oil or district heating. The questions were meant to connect gender to the activities in the house. The stated behavior question was theoretically based on the assumption that stated behavior data has a certain degree of validity (see Whitehead, 2005). In addition, qualitative face-to-face interviews with ten participating households were conducted. The interviews were arranged by calling the telephone number listed for the participants. The participants for the interviews were randomly selected from one municipality within the field study area. The reason for choosing participants from just one municipality was to limit transportation time and increase time spent conducting the interviews. Each of the ten households was interviewed once at the beginning of the field study and once again approximately six months into the field study. The reason for interviewing the households twice was to gain information about their long-term experiences with being part of the project. Each interview lasted 1–1.5 h and took place in the interviewees' homes. In five of the households, both the man and the woman took part in the interviews. In the remaining interviews, only the man took part. In the interviews where only the man took part, the wife was at work. The average age of the interviewees was estimated to be 50–55 years old. During the first round of interviews, the interviewees were asked why they had agreed to participate in the project and what participating in the project had been like to that point. They were also asked to give a tour of the house in order to create a more informal atmosphere and to talk about how they used the different electrical appliances that were in the various rooms and the potential for changing the timing of their electricity consumption for the individual appliances. During the second round of interviews, the interviewees were asked how they felt about participating in the project once again. They were also presented with a weekly time schedule, which they were encouraged to fill out immediately. The schedule was used as the basis for subsequent discussion. The time schedule allowed the interviewees to discuss the (lack of) opportunities they had for changing the timing of their electricity-dependent daily practices. The interviewees were also presented with a 'who does what' form, which they also filled out during the interview. This exercise was used to discuss the different roles of the people in the house and, most importantly, who carried out the electricity consuming practices. If the man and the woman were present during the interview, both of them completed a weekly schedule. However, they were encouraged to fill out the 'who does what' form together in order to stimulate discussion between them. If only the man was present, he filled out the weekly schedule, which was followed by a talk about what the other household members' weekly schedule would have looked like. In the 'who does what' form, he was asked to fill it out on behalf of all the family members. The overall focus of the interviews was the energy consuming practices of the household, the interaction between the household members and the differences between household members regarding who performed the practices. The questionnaire and in-depth qualitative interview methods complimented each other in the sense that the questionnaires reached out to a representative number of people and made it possible to outline general gender tendencies, while the qualitative interviews made it possible to explore ten households in-depth and investigate the context, the reasons for the tendencies outlined in the questionnaires, the interaction between the individuals in the household and to obtain a picture of the practices performed in the participants' own words. The two methods emphasized different theoretical aspects of the concept of gender. The questionnaires were filled out by individuals and focused

Table 2 Overview of information gathering, the questionnaires and the qualitative indepth interviews including duration and key attributes. A. Information gathered before the experiment – Categorization of households. The categories couples and families with children were used for the purpose of this paper – Annual kWh electricity consumption for every household B. Questionnaire at the recruitment (April 2014). The questions used for this study included: – What is your gender and year of birth? – Which cell phone number would you like the text messages sent to? – What is the gender and year of birth of the user of the given cell phone? – Intended behavior question: How important will it be for you to move the timing of your electricity consumption when you, during the test period, receive text messages telling you the best time of the day to use electricity? (1) ❑ Very important (2) ❑ Important (3) ❑ Slightly important (4) ❑ Less important (5) ❑ Not important (6) ❑ Don’t know C. Qualitative in-depth interviews – 1st round of qualitative in-depth interviews with ten households participating in the experiment. The interviews included a tour of the house (August - September 2014). – 2nd round of qualitative in-depth interviews with ten households participating in the experiment. The interviews included filling out a weekly time schedule and a ‘who-does-what’ form (January - February 2015). D. Questionnaire immediately after the end of the experiment (June 2015). The questions used for this study included: – What is your gender and year of birth? – Which of the following appliances do you have in your household (washing machine, dishwasher, tumble dryer, cell phone, PC, etc.)? – Did you (the respondent to the questionnaire) receive the text messages during the experiment? – Stated behavior question: When you received the text messages to change the timing of your electricity consumption, how often did you decide to do the following in order to change the timing of your electricity consumption? Almost Never Don’t Very Often Not never know often very often (1) ❑ (2) ❑ (3) ❑ (4) ❑ (5) ❑ (6) ❑ Changed the timing of washing (1) ❑ Changed the (2) ❑ (3) ❑ (4) ❑ (5) ❑ (6) ❑ timing of dishwashing (1) ❑ (2) ❑ (3) ❑ (4) ❑ (5) ❑ (6) ❑ Changed the timing of tumble drying (1) ❑ (2) ❑ (3) ❑ (4) ❑ (5) ❑ (6) ❑ Changed the timing of charging of cell phone, PC etc.

uncover the participants' pre-existing motivation to increase the flexibility of their electricity consumption. We followed the elicitation format suggested by Sheeran (2002), who presents a comprehensive review of the vast literature on behavioral intentions. The stated intention can indicate people's decision to take particular actions. The aim of the second questionnaire was to determine how frequently the participants had complied with the text messages and had implemented flexible electricity consumption. In the second questionnaire, people were asked to state their behavior. They were asked detailed questions about their practices in the house and how frequently they had

12

Energy Policy 118 (2018) 9–18

L. Tjørring et al.

Fig. 2. Share of people who stated that they very often/often changed the timing of a particular work task when they received text messages (stated behavior after experiment). Fig. 2 is based on following number of responses: washing machine 71 (=100%), tumble dryer 55 (=77%), dishwasher 68 (=94%), and cell phone, PC, etc. 71 (=100%).

Fig. 4. Gender differences in intended behavior and stated behavior about washing. Fig. 4 is based on the following number (N) of responses regarding stated behavior: washing machine N = 71 (=100%), tumble dryer N = 55 (=77%), dishwasher N = 68 (=94%). For the intended behavior question N = 69 responses (=97%). Pearson chi-square tests for difference in compliance rate between women and men receiving text message show that: The stated compliance rates, which were collected after the experiment, are significantly different at the 5%-level for women (compliance rate = 70%, N = 33) and men (compliance rate = 45%, N = 38, chi-square(1, N = 70) = 4.47, p = 0.03). The intended compliance rates, which were collected before the experiment, are not significantly different at the 5% level for women (compliance rate = 81%, N = 32) and for men (compliance rate = 76%, N = 37, chi-square(1, N = 69) = 0.31, p = 0.57). Share of people who stated that it would be very important/important for them to comply with the text messages. Share of people who stated that they had followed the text messages and had very often/often changed the timing of their electricity consumption to the recommended hours. The stated behavior is the average share of people who reported that they very often/often followed the recommendation to change the timing of their electricity consumption to the recommended hours.

Table 3 Extract of results regarding who does what in the ten private households based on a form filled out by the interviewees during the qualitative interview. The form was subsequently discussed with the interviewees. Task

Man

Woman

Both

Laundry House maintenance Cooking Cleaning Doing the dishes

10% 100% 10% 10% 20%

70%

20%

50% 90% 20%

40% 60%

on individual intended and stated behavior. The in-depth qualitative interviews were conducted in order to obtain an understanding of the context of gender in relation to flexible energy consumption from a cultural anthropological perspective.

others. The results are presented in Fig. 2 below. (Table 3) 5.2. When women received the text message, flexible electricity consumption was adopted more frequently

5. Questionnaire results In the questionnaire after the intervention, the households were asked whether it was the man or the woman who had received the text messages. This data was coupled with the participants' answers to the stated behavior question (see Table 1 for full question). Fig. 3 shows the percentage of men and women who said they had complied frequently with the text messages. In the first row, we see that 70% of the women who received text messages said that they had frequently complied with the text messages and had changed the timing of their use of the washing machine, whereas only 42% of the men who received the text messages said that they had frequently complied with the text messages and had changed the timing of their use of the washing machine.

5.1. The households changed the timing of the use of some appliances more than others In the questionnaire after the intervention, the households were asked to state their behavior regarding changing the timing of their electricity consumption (see Table 1 for full question). The questionnaire revealed that for households that very often or often changed the timing of their electricity consumption, the results were 55% for the washing machine, 42% for the tumble dryer, 72% for the dishwasher, and 23% for cell phones, PCs, etc. Thus, the households changed the timing of their use of some electrical appliances more than

Fig. 3. Gender differences in share of people who stated that they very often/often changed the timing of a particular work task when they received text messages (stated behavior after experiment). Fig. 3 is based on the following number (N) of responses: washing machine N = 71 (=100%), tumble dryer N = 55 (=77%), dishwasher N = 68 (=94%), and cell phone, PC, etc. N = 71 (=100%).

13

Energy Policy 118 (2018) 9–18

L. Tjørring et al.

and the 'who does what' form indicated that it was mainly the women who used the washing machine and the tumble dryer, whereas both men and women used the dishwasher. In the majority of the households, the women washed clothes. The women either said that the men did not know how to do this task properly, or they said that they would rather do the laundry than perform other tasks in the house such as maintenance work. Both genders reported that the dishwasher was used more equally. In one household, the man claimed that loading the dishwasher in the most efficient way was a science, which was why he usually did it. In the majority of the households, the man and the woman seemed to agree on their individual responsibilities for running the household: When the respondents were asked to fill out the 'who does what' form, they explained their different roles. The form did not lead to any disagreements in any of the families, and it was completed quickly. An extract from the results of the ‘who does what’ form is presented below. The results of the ‘who does what’ form were similar to the results of the questionnaire. Women predominantly did the laundry in 70% of the households, which involved using the washing machine and the tumble dryer (if they had one). The task of doing the dishes was more equally shared (all households had a dishwasher). In 60% of the households, the man and the woman took turns to do the dishes. In 20% of the households, women predominantly did the dishes, while in the remaining 20% of the households, the men predominantly did the dishes. It should be noted that the sample size was only ten households as the aim was to conduct in-depth qualitative anthropological research rather than achieve representativeness. The questionnaire results indicated the same division of work between man and woman as the qualitative results. Furthermore, domestic gender research also shows this gender division in household tasks. This is discussed in the section below. The interviewees explained that when the woman received the text message, she could immediately act on the instructions within the message because she was responsible for doing the laundry. However, when the man received the text message, he often had to tell his wife what to do. However, if she was not around at the time, he would forget to deliver the message. Because doing the dishes was a shared task in the majority of the households, it did not matter who received the text message. The qualitative results show that when the one 'in charge' of the particular task was also the one who received the text message, it had a significant positive influence on the outcome. In contrast, when the one who was 'not in charge' of the particular task received the text message, an unnecessary middleman was created, which delayed the process/had a negative influence on the outcome.

Significantly more women than men changed the timing of their use of the washing machine7 when they received the text message. In contrast, there was no significant difference between the genders in terms of changing the timing of the use of the dishwasher or cell phone, PC, etc. Thus, the decision to comply varied significantly depending on whether it was a man or a woman who received the text messages. 5.3. Men and women had similar intentions to engage A possible explanation for this difference in behavior is that men and women had different intentions regarding whether to comply with the text messages. Fig. 4 shows the gender difference between intended and stated behavior. The figure is based on the answers to the intended behavior question and the stated behavior question (see Table 2 for full questions). There is no significant difference between the genders in their intention to comply with the text messages. This result is based on households where the same person responded to the questionnaire at the beginning and at the end of the flexible electricity consumption program. In contrast, there is a significant gender difference in the stated behavior. Note that intentions may or may not predict behavior in our sample.8 Our point is that, regardless of whether this is the case or not, other factors could explain the gender difference regarding behavior. We, therefore, consider it relevant to investigate other potential explanations. To sum up, the main result of the questionnaires conducted before and after the intervention is that flexible electricity consumption for the washing machine and the tumble dryer was implemented more frequently when it was a woman who had received the text message. The households also frequently implemented flexible electricity consumption for the dishwasher, but there was no significant difference between the genders concerning this electrical appliance. For the remaining electrical appliances (cell phone, PC, etc.), people implemented flexible electricity consumption less frequently and there was no significant gender difference. The questionnaire results also revealed that there was no significant gender difference in the intention to comply with the text messages, whereas there is a significant gender difference in stated behavior. The results of the questionnaire lead to the following questions: Why is there a gender difference? And why was the timing of electricity consumption changed for some appliances more than others? In order to answer these questions, qualitative in-depth interviews were conducted with ten participating households. 6. Qualitative results

6.2. Why was the timing of electricity consumption changed in some cases, but not in others?

6.1. Why the gender difference? In-depth qualitative interviews were conducted with the participants to investigate their practices in the house. The weekly schedule

During the interviews, the majority of the interviewees said that it was easier to change the timing of their use of the dishwasher than the washing machine or the tumble dryer. The women explained that using the washing machine and the tumble dryer required some planning. First and foremost, the clothes needed to be hung up soon after being washed to avoid them becoming musty. Many people had a tumble dryer, but most carefully considered what clothes to put in. Towels, dishcloths and socks usually went in the tumble dryer, while the female interviewees preferred to hang up more delicate items such as blouses, sweaters and trousers. Therefore, using the washing machine was not an isolated activity, but was connected to other activities (hanging up clothes) that needed to be performed relatively soon afterwards. The majority of women said that they did not have the energy to hang up clothes late in the evening before going to bed, which is why they considered it inconvenient to turn on the washing machine in the evening, even though the electricity was often cheaper. Also, use of the washing machine was connected with sorting clothes according to color and washing temperature, which also affected the flexibility of using

7 Pearson chi-square tests for difference in compliance rate between women and men receiving text message show that: For washing machines, the compliance rate is significantly higher at the 5% level for women (compliance rate = 70%, N = 33) than for men (compliance rate = 42%, N = 38, chi-square (1, N = 71) = 5.43, p = 0.02). For tumble drying, the compliance rates are not significantly different at the 5% level, but close to being significantly different at the 10% level for women (compliance rate = 54%, N = 24) and for men (compliance rate = 32%, N = 31, chi-square(1, N = 54) = 2.66, p = 0.102). For dishwashers, the probabilities of compliance are not significantly different at the 5% level for women (compliance rate = 73%, N = 30), and for men (compliance rate = 71%, N = 38, chi-square(1, N = 67) = 0.04, p = 0.835). For cell phone, etc., the probabilities of compliance are not significantly different at the 5% level for women (compliance rate = 24% N = 33) and for men (compliance rate = 21%, N = 38, chi-square(1, N = 70) = 0.102, and p = 0.748). The probabilities used exclude households without a relevant appliance. 8 The empirical evidence on the connection between intentions and behavior is mixed. Some research finds that there is a connection (Ajzen, 1991; Fishbein, 1980; Triandis, 1980), whereas other studies are more critical (Verplanken et al., 1998; Webb and Sheeran, 2006).

14

Energy Policy 118 (2018) 9–18

L. Tjørring et al.

increased workload as a result of implementing flexible electricity consumption. Rather, the interviewees had relatively fixed considerations and practices concerning which clothes they put in the tumble dryer and which they hung up on the line. Furthermore, participating in the field study did not seem to make the women reduce their use of the tumble dryer. Hence, the workload did not increase; only the time of the day when they performed the tasks changed. What people consider possible to change seems to be connected to cultural norms about acceptable behavior. As the literature in the area points out, there are strong cultural norms connected to washing clothes, which, for example, dictate how often clothes should be washed and at what temperatures (Laermans and Meulders, 2006; Shove, 2003). The female participants in this field study also reported having clear ideas about how to perform the household chores, particularly the laundry. They had a similar systematized approach regarding what clothes needed to be washed at what temperatures, how often clothes needed to be washed and what clothes could go in the tumble dryer and what could not. Other research in the area reports how other countries have different norms for washing clothes. In some countries, it is culturally accepted and considered hygienic to wash clothes in cold water instead of warm (Pakula and Stamminger, 2010), but this is not the case in Denmark. The interviewees found it difficult to change the timing of electricity consumption for the stove, computer and television. An explanation for this is the strong cultural tradition of having dinner at a certain time and in a certain way (Cheng et al., 2007; Powells et al., 2014). Also, the way society is organized with a working day from approximately 8:00–16:00 only leaves time for relaxation including watching TV and using the computer in the evening. Precisely how much change in behavior can realistically be expected is difficult to answer. The concept of 'behavioral plasticity', refers to the percentage of behavioral change that can be realistically expected, based on a review of intervention literature (see Klöckner, 2015:27). Considering this concept, it is found that other empirically-based research classifies 'doing the laundry' as flexible in contrast to activities such as cooking and using the cell phone/PC, which are considered less flexible (Higginson et al., 2014; Powells et al., 2014). In an intervention study from the UK, only 3% of the participating households made use of cheap off-peak electricity (Browne et al., 2013). However, the circumstances of intervention studies always vary, and it is likely that other intervention studies may find higher percentages. Irrespective of the extent of behavioral change, we need to understand activities such as doing the laundry and the dishes in more detail, as has been attempted in this paper, in order to increase the chance of behavioral change in intervention studies. It is important to note that the interviewees had a relatively high average age. Similarly, the respondents of the questionnaire had a high average age of 58 (see Table 1). Detailed information about their work situation (whether they were retired or not) was not gathered. Neither was information about the number of people within each household. However, all households participating in the sub-project about gender within the broader field study were selected to consist of either couples or families with children. Detailed information about the households participating in the qualitative interviews was gathered. Among those ten households, the couple in two of the households was retired. The high average age of 58 may have influenced the results. Previous studies have found a generational effect on energy consumption. Carlsson‐Kanyama et al. (2005) find that older people exhibit more energy efficient behavior than younger people, while Anderson (2016) finds that people with paid work do more laundry during evening peak hours than people without paid work. The results of this paper are limited to apply to people with a relatively high average age. The results of this paper are also limited in the sense that the field study took place in a rural area in Southern Denmark. This may have influenced the results. Other research finds that there is a difference between urban and rural energy consumption. Rural areas are

the washing machine at specific times. In two out of ten households, the man and the woman were both retired. In these households, the women expressed greater flexibility in terms of when it was possible to do the laundry. However, they found it just as hard as the non-retired women to do the laundry late in the evening because of the need to deal with the wet clothes late at night when they were tired. Using the dishwasher was less connected to other activities than using the washing machine. It was not necessary to load or empty it at specific times, which is why both male and female interviewees found it easier to change the timing of their use of the dishwasher to periods with lower electricity prices. Regarding changing the timing of electricity consumption for other appliances, all the interviewees explained that they were so used to having dinner at around 18:00 that they could not change the timing of cooking. The interviewees who had a job explained that because their working day was from approximately 8:00–16:00, it was natural to relax by watching television or using the computer in the evening. The results from the questionnaires and qualitative interviews complimented each other in the sense that the questionnaire indicated that there was a difference between men and women, while the qualitative interviews explained why this difference existed. 7. Discussion of results The focus on families and the interaction between family members showed that flexible electricity consumption was not something that household members reacted to equally. In the case of this field study, the women were the main drivers regarding implementing flexible electricity consumption as they were responsible for the practices, the timing of which could be changed. We need to be aware of the fact that a household often consists of several family members, and flexible electricity consumption is something that affects and is affected by different household members in different ways. The field study showed that this was because household members were responsible for different tasks, which often corresponded with typical gender roles. Hence, the interaction and relationship between household members became important for understanding the potential for implementing flexible electricity consumption. This is in line with the point made by Ellsworth-Krebs et al. (2015); that we need to look at the interaction between household members in order to understand energy consumption. The qualitative results showed that the household members seemed to have a clear agreement about who was responsible for the different tasks and that the women most often did the laundry and cooking. The literature on the division of household tasks confirms that couples who live together have a clear division of labor when it comes to work tasks in the home, and household tasks are heavily gendered (Munro and Madigan, 2006). Whereas women tend to do the cleaning, cooking and laundry, men are primarily responsible for maintaining the house (Carlsson-Kanyama and Lindén, 2007; Tjørring, 2016). The results also showed that it was easier to change the timing of the use of the dishwasher than the washing machine or the tumble dryer because doing the laundry was not an isolated activity, but was connected to other activities such as sorting clothes according to color and temperature and hanging them up to dry. This result corresponds with Warde and Hargreaves' point that tasks are combined, which means that they often co-exist in everyday life and cannot be understood separately (Hargreaves, 2011; Warde, 2005). This is also similar to the findings of Carlsson-Kanyama and Lindén (2007); that when electricity charges vary, the women spend more time doing the laundry as they reduce their use of the tumble dryer, which implies spending more time hanging up clothes on the line. This is also an example of how practices co-exist and how using the washing machine is closely connected to the subsequent drying of clothes. The results of this field study, however, differ from the findings of Carlsson-Kanyama and Lindén (2007) in that the female interviewees did not report an 15

Energy Policy 118 (2018) 9–18

L. Tjørring et al.

having a nice bathroom more important for women than men. Tjørring (2016) also demonstrates that energy renovation falls under the category of maintenance work, which is culturally the man's domain, which explains why men are often more interested in the task. An important point in this regard is that each initiative requires a specific investigation of the practices and persons involved. Of course, gender roles may change over time due to the influence of changes in society. It is possible that men doing the laundry will become a cultural norm in the future and that the practice will change significantly. It is also likely that practices that are now gendered will become less so in the future or vice versa. These changes are difficult to predict, but change is a condition for all intervention strategies whether they focus on gender, economic incentives or technological solutions. Targeting gender in flexible electricity consumption intervention programs also requires ethical considerations. Although flexible electricity consumption can be increased by targeting women instead of men (as this paper suggests), this may have unpredictable side-effects. It may conflict with efforts to achieve gender equality (or the opposite). Being specifically targeted based on a general pattern of gendered division of household tasks (especially among those who do not identify themselves with the general pattern) may also cause harm to women (and men). It should be noted that for this study, women were not specifically targeted. We merely observed that there was a gender difference in the use of electricity based on a subsequent analysis of the data. In fact, social (and gendered) side-effects of intervention studies is a highly understudied area. Skjølsvold et al. (2016) find that smart technologies caused social re-arrangements and friction among household members. Such social (and gendered) side-effects have the potential to emerge in all intervention programs. Focusing on people's gendered practices when designing interventions may have great potential for increasing flexible electricity consumption in private households. In this paper, we have shown that gender has a significant effect on flexible electricity consumption. Using this knowledge will be beneficial to utility companies and public institutions when developing new ways of promoting flexible electricity consumption more effectively. Of course, this may be done in combination with programs that generally introduce, for example, economic incentives or specific technological solutions. What is special about a focus on gendered practices compared to economic incentives and technological solutions is that it addresses the problem of excessive and inflexible electricity consumption from an everyday gendered practice perspective. Most private electricity consumption occurs as a result of everyday practices in the home. Understanding this (gendered) everyday level of consumption is crucial for developing long-term solutions. However, more research is needed to determine how a gender perspective can be practically included in interventions and energy policy.

characterized by a need for energy-intensive private transport and the heating of larger houses, whereas people in urban areas tend to have higher total energy requirements (Lenzen, 1998; Wiedenhofer et al., 2013). Regarding gender, older people living in rural areas may have more distinct gender roles than younger people who live in the city. However, Denmark ranks second on the international gender equality index (European Institute for Gender Equality, 2013), and it is likely that the extent of the gender differences found in this rural sample is similar to, or more pronounced in, other countries. To sum up, the study results may differ in city areas in terms of both gender differences and energy requirements. It is also important to note that the questionnaire results are based on self-reported data and are, therefore, prone to inaccurate recall and error. However, we have no reason to believe that there is a systematic bias (i.e. that men and women report systematically differently). Furthermore, a review of the validity of self-reported measures indicates a general association between self-reported measures of proenvironmental behavior and objective measures (Kormos and Gifford, 2014). With regards to the future, in order to accommodate more renewable energy production in the grid, in Denmark, there is a focus on converting to electricity-powered appliances. For example, replacing oil-burners with heat pumps and promoting electrical vehicles rather than vehicles powered by fossil fuels. In this paper, we have not focused on heating or transport as, in general, these are not powered by electricity. However, some research with a focus on heating and transport is slowly emerging in a Danish context (see, for example, Friis and Haunstrup Christensen, 2016; Haunstrup Christensen et al., 2013; Nyborg and Røpke, 2013). To sum up, we set out to investigate the connection between the implementation of flexible electricity consumption and gender in families that consist of at least two family members. We considered gender to be a cultural concept and one which is closely related to the practices performed in the household. The study found that the timing of the use of energy consuming appliances, such as the washing machine, dishwasher and tumble dryer could be changed to different times of the day. Besides, doing the laundry was mainly performed by women, which was why the implementation of flexible electricity consumption increased when women received the text message from the energy company. In this section, we have highlighted how gender roles in the household and the practices to which they are attached are interconnected with cultural norms concerning not just who performs the practices, but also how they are performed, how they are connected to other practices and when it is possible to perform them. 8. How to practically include a gender focus when designing initiatives for flexible electricity consumption In order to practically incorporate gender when promoting flexible electricity consumption, we need to answer the question "who performs the practice we want to change"? In this field study, text messages directed at women had a positive effect on the outcome because the women were responsible for the particular practices we wanted to change. It is likely that other ways of addressing women would have had a similar effect (for example, placing meters in women's parts of the house, or directing campaigns at women). Energy-related practices range from renovating houses to improving energy efficiency to using an electrical vehicle or operating an automated smart home. In this connection, it is likely that gender roles are at play regarding the different practices. For example, Tjørring (2016) shows that the decision to renovate a house in order to improve energy efficiency and what sort of energy improvements to make are closely related to the different practices men and women perform in the house. Men and women use the house in different ways and, hence, have different preferences in terms of what to do with the house. For example, women often use the bathroom more than men, which makes

9. Conclusion and policy recommendations Taking gendered practices into account has the potential to increase the effectiveness of policies which aim to change the timing of energy consumption in private households by introducing flexible electricity prices. In this paper, we have demonstrated that energy-consuming household practices are gendered. Furthermore, there was a significant difference in stated behavior responses regarding changing the timing of energy consumption depending on whether it was a man or a woman who had received the text message. The study showed that the correct strategy was to address the woman in the household as she was normally solely or partially responsible for using the relevant energy-consuming appliances (washing machine, tumble dryer, etc.) and could take action and change the timing of electricity consumption relatively quickly. When using text messages as in this study, it is important to send the message to the person who is in charge so that they can then take action. The process was hampered when a person who was not in charge received the 16

Energy Policy 118 (2018) 9–18

L. Tjørring et al.

Document]. URL 〈http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?Uri= COM:2011:0202:FIN:EN:PDF〉 (Accessed 14 April 2016). European Environment Agency, 2015. Key assessment [WWW Document]. URL 〈http:// www.eea.europa.eu/data-and-maps/indicators/household-energy-consumption〉 (Accessed 28 December 2015). European Institute for Gender Equality, 2013. Gender Equality Index Report [WWW Document]. URL 〈http://eige.europa.eu/sites/default/files/documents/ mh0215616enn.pdf〉 (Accessed 28 December 2015). Fishbein, M., 1980. A theory of reasoned action: some applications and implications. In: Howe, H., Page, M. (Eds.), Nebraska Symposium on Motivation. University of Nebraska Press, Lincoln, pp. 65–116. Friis, F., Haunstrup Christensen, T., 2016. The challenge of time shifting energy demand practices: insights from Denmark. Energy Res. Soc. Sci. 19, 124–133. Hargreaves, T., 2011. Practice-ing behaviour change: applying social practice theory to pro-environmental behaviour change. J. Consum. Cult. 11, 79–99. Haunstrup Christensen, T., Throndsen, W., Ascarza, A., Gram-Hanssen, K., Friis, F., 2013. The role of households in the smart grid: a comparative study. ECEEE Summer Study Rethink. Renew. Restart Dyn. Consum. 2285–2296. Higginson, S., Thomson, M., Bhamra, T., 2014. “For the times they are a-changin”: the impact of shifting energy-use practices in time and space. Local Environ. 19, 520–538. Isaksson, C., Ellegård, K., 2015. Dividing or sharing? A time-geographical examination of eating, labour, and energy consumption in Sweden. Energy Res. Soc. Sci. 10, 180–191. Jensen, C.L., Hansen, L.G., Fjordbak, T., Gudbjerg, E., 2012. Providing free autopoweroff plugs: measuring the effect on households’ electricity consumption through a field experiment. Energy J. 33, 187–211. Karki, R., Billinton, R., 2001. Reliability/cost implications of PV and wind energy utilization in small isolated power systems. IEEE Trans. Energy Convers. 16, 368–373. Kempton, W., Layne, L.L., 1994. The consumer's energy analysis environment. Energy Policy 22, 857–866. Klöckner, C.A., 2015. The Psychology of Pro-environmental Communication: Beyond Standard Information Strategies, The Psychology of Pro-Environmental Communication: Beyond Standard Information Strategies. Palgrave Macmillan, New York. Kormos, C., Gifford, R., 2014. The validity of self-report measures of proenvironmental behavior: ameta-analytic review. J. Environ. Psychol. 40, 359–371. Laermans, R., Meulders, C., 2006. The domestication of laundering. In: Cieraad, I. (Ed.), At Home: An Anthropology of Domestic Space. Syracuse University Press, New York, pp. 118–130. Lenzen, M., 1998. Energy and greenhouse gas cost of living for Australia during 1993/94. Energy 23, 497–516. Lutzenhiser, L., 1992. A cultural model of household energy consumption. Energy 17, 47–60. Lutzenhiser, L., Shove, E., 1999. Contracting knowledge: the organizational limits to interdisciplinary energy efficiency research and development in the US and the UK. Energy Policy 27, 217–227. Munro, M., Madigan, R., 2006. Negotiating space in the family home. In: Cieraad, I. (Ed.), At Home: An Anthropology of Domestic Space. Syracuse University Press, New York, pp. 107–118. Nyborg, S., Røpke, I., 2013. Constructing users in the smart grid-insights from the Danish eFlex project. Energy Effic. 6, 655–670. Pakula, C., Stamminger, R., 2010. Electricity and water consumption for laundry washing by washing machine worldwide. Energy Effic. 3, 365–382. Pink, S., 2012. Situating Everyday Life: Practices and Places. Sage Publications, London. Powells, G., Bulkeley, H., Bell, S., Judson, E., 2014. Peak electricity demand and the flexibility of everyday life. Geoforum 55, 43–52. Ryan, S.E., 2014. Rethinking gender and identity in energy studies. Energy Res. Soc. Sci. 1, 96–105. Räty, R., Carlsson-Kanyama, A., 2010. Energy consumption by gender in some European countries. Energy Policy 38, 646–649. Sayer, L.C., 2010. Trends in housework. In: Treas, J., Drobnic, S. (Eds.), Dividing the Domestic: Men, Women, and Household Work in Cross-National Perspective. Stanford University Press, Stanford, pp. 19–41. Sheeran, P., 2002. Intention—Behavior relations: a conceptual and empirical review. Eur. Rev. Soc. Psychol. 12, 1–36. Shove, E., 2003. Comfort, Cleanliness and Convenience: The Social Organization of Normality. Berg, Oxford. Skjølsvold, T.M., Jørgensen, S., Ryghaug, M., 2016. Users, design and the role of feedback technologies in the Norwegian energy transition: An empirical study and some radical challenges. Energy Res. Soc. Sci. 25, 1–8. Sovacool, B.S., Saleem, S., D’Agostino, A.L., Ramos, C.R., Trott, K., Ong, Y., 2012. What about social science and interdisciplinarity? A 10-year content analysis of energy policy. In: Goldblatt, D., Minsch, J., Flüeler, T., Spreng, D. (Eds.), Tackling Long-Term Global Energy Problems. Springer, New York, pp. 47–71. Strengers, Y., 2012. Peak electricity demand and social practice theories: reframing the role of change agents in the energy sector. Energy Policy 44, 226–234. The Danish Government, 2011. Energy Strategy 2050 - From Coal, Oil and Gas to Green Energy [WWW Document]. URL 〈http://www.stm.dk/multimedia/Energistrategi_ 2050.pdf〉 (Accessed 29 May 2017). The Government of Norway, 2012. Meld. St. 21 (2011–2012) [WWW Document]. URL 〈https://www.regjeringen.no/no/dokumenter/meld-st-21-2011-2012/id679374/ ?Ch=1&q=〉 (Accessed 28 December 2015). The Government of Sweden, 2008. En sammanhållen klimat- och energipolitik [WWW Document]. URL http://www.regeringen.se/contentassets/cf41d449d2a047049d7a34f0e23539ee/en-sammanhallen-klimat–och-energipolitik—klimat-prop.200809162 (Accessed 28 December 2015). Tjørring, L., 2016. We forgot half of the population! The significance of gender in Danish energy renovation projects. Energy Res. Soc. Sci. 22, 115–124. Triandis, H.C., 1980. Values, attitudes, and interpersonal behavior. In: Howe, H., Page, M.

message. On a more general level, this paper indicates a need to ask the question 'who performs the practice we want to change?', when promoting flexible electricity consumption. There is a range of energyconsuming practices, and they most likely imply different gender divisions. We need to establish a more direct link between energy policies and gender, so that energy policies can be developed which take the important issue of gender into account. We suggest that this would require an analysis of the gender roles at play, the practices to which they are attached and the cultural norms that form these gender roles and the performance of household tasks. However, this points to the need for more research that focuses on how to practically include a gender focus in new policy programs and how to be sensitive to the different contexts in which different gender roles are at play. Acknowledgements We would like to express our gratitude to Innovation fund Denmark, the energy company SE and the University of Copenhagen for funding the research upon which this paper is based, and to the SE energy company for making it possible to conduct the field study. We would like to thank the individuals who participated in the field study for completing the questionnaires and the ten households for putting time aside to participate in the qualitative interviews. Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.enpol.2018.03.006. References Ajzen, I., 1991. The theory of planned behavior. Organ. Behav. Hum. Decis. Process. 50, 179–211. Altinget, 2016. Her er regeringens energikommision [WWW Document]. URL 〈https:// www.altinget.dk/artikel/her-er-energikommissionens-medlemmer〉 (Accessed 14 April 2016). Anderson, B., 2016. Laundry, energy and time: insights from 20 years of time-use diary data in the United Kingdom. Energy Res. Soc. Sci. 22, 125–136. Aune, M., 2007. Energy comes home. Energy Policy 35, 5457–5465. Bianchi, S.M., Milkie, M.A., Sayer, L.C., Robinson, J.P., 2000. Is anyone doing the housework? Trends in the gender division of household labor. Soc. Forces 79, 191–228. Browne, A.L., Medd, W., Anderson, B., 2013. Developing novel approaches to tracking domestic water demand under uncertainty-A reflection on the “Up scaling” of social science approaches in the United Kingdom. Water Resour. Manag. 27, 1013–1035. Carlsson-Kanyama, A., Juliá, I.R., Röhr, U., 2010. Unequal representation of women and men in energy company boards and management groups: are there implications for mitigation? Energy Policy 38, 4737–4740. Carlsson-Kanyama, A., Lindén, A.-L., 2007. Energy efficiency in residences—Challenges for women and men in the North. Energy Policy 35, 2163–2172. Carlsson‐Kanyama, A., Lindén, A., Eriksson, B., 2005. Residential energy behaviour: does generation matter? Int. J. Consum. Stud. 29, 239–253. Cheng, S., Olsen, W., Southerton, D., Warde, A., 2007. The changing practice of eating: evidence from UK time diaries, 1975 and 2000. Br. J. Sociol. 58, 39–61. Chu, S., Majumdar, A., 2012. Opportunities and challenges for a sustainable energy future. Nature 488, 294–303. Clancy, J., Roehr, U., 2003. Gender and energy: is there a Northern perspective? Energy Sustain. Dev. 7, 44–49. Committee on Climate Change, 2015. The Climate Change Act and UK Regulations [WWW Document]. URL 〈https://www.theccc.org.uk/tackling-climate-change/thelegal-landscape/global-action-on-climate-change/〉 (Accessed 28 December 2015). D’Agostino, A.L., Sovacool, B.K., Trott, K., Ramos, C.R., Saleem, S., Ong, Y., 2011. What's the state of energy studies research?: A content analysis of three leading journals from 1999 to 2008. Energy 36, 508–519. Danish Energy Agency, 2016. Energistatistik 2015 [WWW Document]. URL 〈https://ens. dk/sites/ens.dk/files/Statistik/energistatistik2015.pdf〉 (Accessed 15 May 2017). Danish Energy Association, Energinet.dk, 2013. Smart grid i Danmark 2.0. København. Dietz, T., Kalof, L., Stern, P.C., 2002. Gender, values, and environmentalism. Soc. Sci. Q. 83, 353–364. Ellegård, K., Palm, J., 2015. Who is behaving? Consequences for energy policy of concept confusion. Energies 8, 7618–7637. Ellegård, K., Palm, J., 2011. Visualizing energy consumption activities as a tool for making everyday life more sustainable. Appl. Energy 88, 1920–1926. Ellsworth-Krebs, K., Reid, L., Hunter, C.J., 2015. Home-ing in on domestic energy research:“house,”“home,” and the importance of ontology. Energy Res. Soc. Sci. 6, 100–108. European Commision, 2011. Smart Grids: From Innovation to deployment [WWW

17

Energy Policy 118 (2018) 9–18

L. Tjørring et al.

(Eds.), Nebraska Symposium on Motivation. University of Nebraska Press, Lincoln NB, pp. 195–259. Verplanken, B., Aarts, H., Knippenberg, A., Moonen, A., 1998. Habit versus planned behaviour: a field experiment. Br. J. Soc. Psychol. 37, 111–128. Walker, G., 2014. The dynamics of energy demand: change, rhythm and synchronicity. Energy Res. Soc. Sci. 1, 49–55. Warde, A., 2005. Consumption and theories of practice. J. Consum. Cult. 5, 131–153. Webb, T.L., Sheeran, P., 2006. Does changing behavioral intentions engender behavior change? A meta-analysis of the experimental evidence. Psychol. Bull. 132, 249–268. West, C., Zimmerman, D.H., 1987. Doing gender. Gend. Soc. 1, 125–151. Whitehead, J.C., 2005. Environmental risk and averting behavior: predictive validity of jointly estimated revealed and stated behavior data. Environ. Resour. Econ. 32, 301–316.

Widén, J., Lundh, M., Vassileva, I., Dahlquist, E., Ellegård, K., Wäckelgård, E., 2009. Constructing load profiles for household electricity and hot water from time-use dataModelling approach and validation. Energy Build. 41, 753–768. Wiedenhofer, D., Lenzen, M., Steinberger, J.K., 2013. Energy requirements of consumption: urban form, climatic and socio-economic factors, rebounds and their policy implications. Energy Policy 63, 696–707. Wilhite, H., Shove, E., Lutzenhiser, L., Kempton, W., 2000. The legacy of twenty years of energy demand management: we know more about individual behaviour but next to nothing about demand. In: Jochem, E. (Ed.), Society, Behaviour, and Climate Change Mitigation. Kluwer Academic Publishers, New York, pp. 109–126. Zelezny, L.C., Chua, P., Aldrich, C., 2000. New ways of thinking about environmentalism: elaborating on gender differences in environmentalism. J. Soc. Issues 56, 443–457.

18