Abstract
In the last decade, South Africa has frequently experienced electricity shortages. Conserving electricity is a sustainable means by which households can contribute to mitigating the problem. The main electricity supplier, Eskom, partners with the South African government to promote electricity conservation, particularly in households. For effective tailoring of promotional messages, market segmentation is needed. This study uses cluster analysis to segment and identify segment sizes and types of electricity conservation behaviors in South African households. It also profiles the segments according to sociodemographic characteristics and some economic and psychological drivers of conservation behaviors. We surveyed 629 electricity consumers in the Gauteng province of South Africa. Four segments were identified: devoted conservers (18%), unconcerned nonconservers (28.7%), curtailment-oriented conservers (29.3%), and efficiency-oriented consumers (24%). The article proposes germane strategies to be tailored by social marketers and policy makers to reach each of these segments.
Keywords
Electricity is undoubtedly fundamental for the social and economic growth of any country. Regrettably, South Africa has frequently experienced electricity shortages in the past decade, caused by increased demand from a rising population, urbanization, and industrialization. In periods when demand exceeds supply capacity, there are blackouts, causing businesses to incur losses and many households to be left for days without electricity (International Energy Agency, 2011). To solve the electricity shortage problem on the supply side, new power stations are being built.
On the demand side, various media, including the country’s national TV, have been widely used, not only to create awareness about the electricity shortages but also to foster efficient electricity consumption habits. Prominent among the campaigns was the 49M initiative, which called on all 49 million South Africans, during the campaign period in 2011, to avoid the unrestrained use of electricity (Eskom, 2011). While these efforts substantially increased public awareness of the electricity shortage problem and suggested saving measures, the actual electricity saving or conserving behavior did not significantly change (South African Department of Energy [DoE], 2013). This suggests that the barriers to electricity conservation remain behavior-specific.
Recent developments indicate that the national electric grid was stabilized when additional supply power stations went operational in October 2015 (Eskom, 2016a). While increased supply from the stations is a welcome relief, electricity shortfalls are still imminent, given the fast-growing population and increased business activities (Asif & Muneer, 2007). In addition, the increased usage of coal in electricity production stations creates environmental problems. A more sustainable solution is therefore a demand-based one that entails radical behavior change toward parsimonious electricity consumption. Social marketing approaches have proved successful in enhancing energy conservation in households (Gray & Bean, 2011; Reaves, Clevenger, Nobe, & Aloise-Young, 2016; Schultz et al., 2015); however, according to Newton, Newton, Turk, and Ewing (2013), market segmentation should be the starting point for a better understanding of the targeted market, better tailoring of offerings, repositioning of behavior, and more effective promotion of behavior change.
The usefulness of market segmentation in fostering pro-environmental behaviors within the social marketing framework has been widely acknowledged (Gray & Bean, 2011; Lotenberg, Schechter, & Strand, 2011). An enquiry into extant literature on market segmentation of energy efficiency (Sütterlin, Brunner, & Siegrist, 2011) and pro-environmental behavior as a whole (Barber, Bishop, & Gruen, 2014; Barr, Gilg, & Ford, 2005; Barr, Shaw, & Coles, 2011) shows a growing interest in the subject.
However, some of these environmental studies have merely divided the markets into groups based on the level of commitment to environmental behaviors (e.g., nonconservers, low conservers, and committed conservers) and thus failed to provide additional psychological motives behind this behavior (e.g., Thomas, 2001). Other studies have provided a comprehensive segmentation basis that encompasses past behaviors, sociopsychological and financial motives, and demographical characteristics (e.g., Barr et al., 2011; Sütterlin et al., 2011). However, these studies addressed only the barriers and motivations for adopting energy conservation in general. Some dealt with pro-environmental behaviors broadly, which comprised a wide range of environmental actions, such as energy conservation, recycling, and water conservation. Yet even in energy conservation, there are many types of energy (e.g., fuel, electricity), consumed in households, businesses, or public spaces. Broadly segmenting pro-environmental behavior may therefore not be useful to understanding a specific conservation behavior (McKenzie-Mohr, Lee, Schultz, & Kotler, 2012; McKenzie-Mohr & Schultz, 2014). According to McKenzie-Mohr, Lee, Schultz, and Kotler (2012) and Segev (2015), an effective social marketing initiative should be focused on a specific conservation behavior (e.g., switching off lights to save electricity or taking shorter showers) within a specific sector (e.g., residential).
Moreover, the extant energy conservation literature contains little market segmentation research in the field of social marketing that has been conducted in a developing multicultural country such as South Africa and used a rigorous statistical technique like cluster analysis. Based on the identified research gaps in literature, this study uses cluster analysis to segment South African electricity consumers in the residential sector on the basis of their past electricity saving actions. It further profiles these segments according to underpinning psychological and financial motives as well as sociodemographic characteristics. The market segmentation and identification of electricity conservation behavioral types could not only guide policy makers and the tailoring of appropriate and effective social marketing campaigns but also promote sustainable electricity conservation behaviors in South African households.
The remainder of this article outlines the importance of market segmentation in the social marketing process of behavioral change, highlights important findings of the previous studies, and discusses the methodology and results obtained in this study.
Rationale for Segmenting Residential Electricity Consumers
The residential sector plays a fundamental role in escalating electricity demand in many countries. In South Africa, for example, households consume about 20% of electricity and contribute to more than 30% of peak period demand, which has often led to blackouts (Eskom, 2016b). It has been demonstrated that if all South African households could reduce their electricity consumption by 10%, it would be as effective as building a new power station and much cheaper (Eskom, 2011). Reducing household electricity consumption is a more sustainable and less costly option for keeping the lights on, but identifying the group of consumers to best target with this message requires market segmentation and profiling.
Identification of the groups expected to change behavior is also important for appropriate social marketing campaigns (Barr et al., 2011; Gray & Bean, 2011). Market segmentation is also essential, because similarities exist in consumers’ needs and their perception of behavior changes across demographic, psychological, and social groupings (Lotenberg et al., 2011; Newton, Newton, Turk, & Ewing, 2013). A study by Gray and Bean (2011) found segmentation in social marketing to be an important step in bringing about effective electricity conservation in Australian households. Market segmentation, the authors explain, provides greater insight into electricity consumers’ specific issues and behavioral patterns, which in turn leads to the provision of tailored information and targeted social marketing campaigns for behavior change. To learn how best to conduct such market segmentation, it is important to review prior literature on market segmentation, even though it has broadly focused on pro-environmental behaviors.
Previous Studies on Segmenting and Profiling the Conservation Behavior of Energy Consumers
There have been several attempts to segment and profile pro-environmental consumers (Barber et al., 2014; Barr et al., 2005; Vicente & Reis, 2007). Studies have generally described pro-environmental consumers as educated women, premiddle aged, with substantial incomes (Barr et al., 2005; Finisterra do Paço & Raposo, 2010; Sütterlin et al., 2011; Vicente & Reis, 2007). From reviewing energy conservation literature, as Sütterlin, Brunner, and Siegrist (2011) regretfully observe, it appears that few market segmentation studies have used advanced statistical techniques like cluster analysis, which better differentiates and provides an accurate picture of energy consumer types.
When segmenting and profiling consumers, studies such as those by Barr, Gilg, and Ford (2005), Finisterra do Paço and Raposo (2010), and Jansson, Marell, and Nordlund (2009) have considered pro-environmental behaviors as one single behavior, without differentiating and isolating the specific pro-environmental actions that various segments undertake. For example, Barr, Shaw, and Coles’s (2011) segmentation of UK pro-environmentalists revealed that there are three segments. In Segment 1, consumers are mostly eco-conscious individuals who are very committed, frequently undertake home-based environmental actions and purchase environmentally friendly and organic products. Segment 2 consumers are strongly committed to recycling, composting, and doing energy conservation activities at home but are less committed in purchasing energy-saving products. Segment 3 consumers participate in recycling and are very keen on conserving water and energy but do not engage in any other forms of environmental behavior, such as purchasing environmentally friendly products. Barr et al. (2011) further found that the environmental behaviors of their identified three clusters are heavily contextualized by the sites in which the behaviors are performed. This raises the question of the effectiveness of positive “spillover effect,” which assumes that the implementation of one pro-environmental behavior (e.g., electricity conservation) can snowball to the implementation of other activities (e.g., recycling; Thøgersen & Crompton, 2009).
A study undertaken in 2007 by the Australian DoE Conservation of New South Wales segmented Australian consumers according to wide-ranging pro-environmental behaviors. The segmentation was based on citizenship behaviors (e.g., encouraging community members to adopt pro-environmental actions), purchasing behaviors (e.g., avoiding using plastic bags in order to protect the environment), and actions in households (e.g., electricity conservation, recycling, and water conservation). This comprehensive market segmentation identified four distinctive segments: “committeds,” “middles,” “privates,” and “reluctants” (Australian Department of Environment and Conservation, 2007).
Another comprehensive market segmentation is that of the Swiss energy market conducted by Sütterlin et al. (2011). Their cluster analysis revealed six distinct segments comprising the “idealistic,” “selfless inconsequent,” “thrifty,” “materialistic,” “convenience-oriented indifferent,” and “problem-aware well-being-oriented.” These six energy consumer types are distinguished according to their level of adoption of energy-saving measures in the housing, food, and mobility domains (past behaviors); their degree of acceptance of policy measures in the nuclear, sales, and mobility domains; and some energy-related sociopsychological and economic motives. For example, the “selfless inconsequent” energy consumer type is likely to accept energy-saving policy regulations, such as the increase in price of cars with high energy consumption, but puts less effort into the actual implementation of energy conservation actions, such as the use of public transport or a bicycle to save energy. The “thrifty” energy consumer type engages in energy-saving actions due to financial considerations but is less likely to accept energy-saving regulations.
While previous studies have profiled pro-environmental consumers in a comprehensive way (Barr et al., 2005; Sütterlin et al., 2011; Vicente & Reis, 2007), it is questionable whether the same precise profile and size of segments can be found in the electricity sector, particularly in a culturally and socioeconomically diverse country like South Africa. For example, the “idealistic energy-saver” segment, which Sütterlin et al. (2011) found in Switzerland, consisting of those consumers who engage most in energy-conserving efforts, comprised up to 15.5%. It is doubtful whether the “idealistic energy-saver” segment will be found in South Africa, given that in terms of Hofstede’s cultural dimension model, South Africa scores fairly high (63) for “indulgence” (the general willingness of people to fulfill their impulses and desires to have fun and please themselves as they wish), which is higher than most European states (The Hofstede Centre, 2014). In addition, Segev (2015) suggests that conservation behavior differs in terms of the type of activity concerned (e.g., water and electricity conservation and recycling) and the cultural values the consumers hold.
Similarly, Thøgersen and Crompton (2009) and Steinhorst, Klöckner, and Matthies (2015) are concerned that the adoption of a pro-environmental behavior has not been adequately proved to spillover to other such behaviors. This has left questions about the generalizations that can be made from previous pro-environmental behavior segmentation. Whether the segment profiles found in other studies will apply when electricity conservation behavior is examined has been questioned.
The few studies which have segmented and profiled electricity conservation behavior and used behavioral segmentation bases have distinguished between two dimensions of electricity-conserving behaviors, namely, curtailment and efficiency actions (e.g., Barr et al., 2005; Sütterlin et al., 2011). Curtailment actions refer to recurrent and/or low-cost (or free) energy-saving behaviors, which entail a cutback on amenities or comfort and must be repeated (Karlin et al., 2014). Examples of these actions include unplugging home appliances, switching off lights when leaving the room, and using less hot water for the shower or bath.
In contrast, efficiency behaviors are infrequent structural changes and/or involve financial investments or purchases but result in no or very little loss of comforts and have a longer lasting energy conservation effect (Karlin et al., 2014). An example of such action is installing energy efficient products or equipment, such as roof and wall insulations. These behaviors are driven by several factors discussed in the next section.
Commonly Found Determinants of Energy Conservation Behavior Used in the Current Study
A growing body of research has been devoted to determining the major factors that shape residential energy conservation (Frederiks, Stenner, & Hobman, 2015; Steg & Vlek, 2009; Thøgersen & Grønhøj, 2010). The key findings have been categorized into four main determinants: psychological motives, sociodemographic characteristics, contextual variables (e.g., financial or economic standing), and habits (Steg & Vlek, 2009; Thøgersen & Grønhøj, 2010). Given that sociopsychological, economic, and sociodemographic constructs have been proved to have high explanatory power for energy conservation (Finisterra do Paço & Raposo, 2010; Mostafa, 2009), these drivers were used to profile electricity consumers in this study.
Psychological motives underpinning energy conservation, developed in the extant literature, are mainly drawn from the theory of planned behavior (Ajzen, 1991) and the norm activation model (Schwartz & Howard, 1981). Following previous research (e.g., Barr et al., 2005; Diaz-Rainey & Ashton, 2011; Sütterlin et al., 2011), which particularly found the psychological constructs of attitude, social norms, personal norms, ascription of responsibility, feelings of guilt, and behavioral intention to be significant drivers of energy conservation, these constructs were used to profile electricity consumers in the present study.
Attitude toward saving electricity refers to the overall positive or negative evaluation of the electricity conservation actions (Maio & Haddock, 2010). Social norms (or subjective norms) refer to the perception that important others are or are not saving electricity (Fishbein & Ajzen, 2010). Personal norms (or moral norms) in this study refer to the internal or moral obligation to conserve electricity (Steg & De Groot, 2010). Feelings of guilt translate to a self-conscious emotion reflecting one’s response to personal responsibility for saving or not saving enough electricity (Harth, Leach, & Kessler, 2013). Ascription of responsibility refers to feelings of personal responsibility for the adverse consequences of not using electricity parsimoniously (Steg & De Groot, 2010).
The common demographic characteristics found relevant to electricity conservation and used as profiling constructs for the current study are location (township or suburb), type of house, electric meter, homeownership, gender, age-group, educational level, racial group, monthly gross income, and monthly expenditure on electricity.
Incentives and electricity price increases are two economic drivers that have an impact on people’s decision to conserve energy (Kollmuss & Agyeman, 2002; Steinhorst, Klöckner, & Matthies, 2015). To this end, the inclining block tariff (IBT) is a punitive electricity pricing strategy which discourages high electricity consumption in South Africa (DoE, 2013). The more kilowatt-hour a billed consumer uses, the higher the average charged price.
Research Methodology
Sample and Data Collection
This article follows a single cross-sectional design. A nonprobability, quota sampling method was used. The use of quota sampling methods, instead of uncertain and sometimes costly probability sampling methods, is often recommended in social marketing studies (Kotler & Lee, 2008). The quotas were based on the respondent’s gender, racial group, and geolocation types (suburban and nonurban location), reflected by South Africa’s national census (Statistics South Africa, 2014). The sample unit consisted of heads of households living in the Gauteng province in both the suburban and the nonurban (township) areas. The choice of Gauteng province was motivated by its eclectic population, vast economic activities, and the province’s record of the highest electricity consumption in the country (Musango, 2014).
The questionnaire consisted of three sections. The first section featured the demographic information of respondents and household characteristics. The second was geared toward assessing the curtailment and efficiency dimensions of electricity conservation in households. The third section measured the economic and psychological determinants of electricity conservation actions. All the scales were adapted from previous studies (e.g., Abrahamse & Steg, 2009; Botetzagias, Malesios, & Poulou, 2014; Sütterlin et al., 2011; Thøgersen & Grønhøj, 2010). The electricity-saving measures were obtained from available guides describing how to conserve electricity in the South African context (e.g., 49 Million, 2015).
Data were collected by means of self-administered questionnaires from May to September 2015. Paper-based and online versions of the questionnaires were administered to increase their reach in terms of contacting heads of households. The hyperlink for the online questionnaire was e-mailed to the university academic and support staff. This was because the researchers were working at the university, and the e-mail addresses of the respondents could be accessed. The number of usable questionnaires received from the online survey was 159. The paper-based questionnaires were widely distributed to other heads of households inside and outside of the university, in cafes, churches, and malls and with permissions from the authorities. A total of 610 paper-based questionnaires were received from the 800 distributed, with a response rate of 76%. Of these, 470 were found usable after cleaning the data set, removing outliers, and excluding cases exceeding the predefined quotas. The sum of the usable questionnaires finally obtained from both the online and paper-based surveys, and included in the data analysis, was 629.
Data Analysis and Results
Data Analysis Process
The data were analyzed with the SPSS 22 package. Segments were first identified, based on past electricity curtailment and efficiency actions. Thereafter, the segments were profiled on the basis of the selected psychological, economic, and sociodemographic determinants of electricity conservation.
Given that the scales used for this study were adapted from previous studies, a confirmatory factor analysis (CFA) was conducted to confirm the structure of the scales (Hair, Black, Babin, & Anderson, 2014). The CFA was performed in AMOS Version 22 (Table 1). All the loading estimates were highly significant. For greater variance, only items with a factor loading above 0.3 were considered (Field, 2013).
CFA Results, Mean, and Reliability αs of This Study’s Variables.
Note. CFA = confirmatory factor analysis.
aThe construct electricity conservation by efficiency is measured by dichotomous yes/no questions. Yes is codified as 1 and No as 0. The electricity efficiency index ranges from 0 to 8 (total number of items), which indicates the number of implemented electricity efficiency measures.
Reliability Test and Descriptive Statistics
The reliability of each scale in terms of the internal consistency was measured through Cronbach’s α. For almost all the scales, Cronbach’s α coefficients were above the threshold of .7, which is usually considered acceptable for internal consistency of the scale (Hair et al., 2014). Only the electricity conservation by curtailment (α = .683) and efficiency (α = .609) variables displayed coefficients lower than .7. This is consistent with other energy conservation studies (e.g., Abrahamse & Steg, 2009; Thøgersen & Grønhøj, 2010), which obtained these close to acceptable coefficients, probably because the scale measures a variety of actions toward electricity conservation.
Cluster Analysis
A cluster analysis of the data was performed to identify the segments. The assumption of multicollinearity among clustering variables was assessed through the variance inflation factor (VIF). All VIFs were well below the cutoff of 10, indicating that there is no effect of multicollinearity. The assumption was therefore met. Following Hair, Black, Babin, and Anderson’s (2014) clustering approach, a hierarchical cluster analysis built on standardized variables was first conducted to ascertain the optimal number of clusters. Ward’s method and the square Euclidean distance as proximity measure were applied to determine an initial cluster solution. An analysis of the agglomeration schedule, combined with a close observation of the dendrogram, denoted a four-cluster solution. On account of this result, a k-means cluster analysis, which is a nonhierarchical clustering technique that provides additional elaborative information about the clusters, was performed in the second step. The clustering variables were standardized and the k-means analysis was performed, based on the four-cluster solution. A convergence was achieved after six iterations. A scrutiny of the value of the cluster center, size of the cluster, and distance between clusters confirmed the four-cluster solution (Table 2). The cluster membership was saved as a new categorical variable in the data set. From a one-way analysis of variance (ANOVA) test, significant differences were observed across clusters for the curtailment (F = 446.416; p < .001) and efficiency conservation behaviors (F = 465, .587; p < .001).
Cluster Centers of Continuous Variables.
Note. The constructs are represented by their standardized Z-scores. Monthly expenditure shows the real value in Rand. 1 Rand = US$14. Cluster 1 = devoted conservers; Cluster 2 = unconcerned nonconservers; Cluster 3 = curtailment-oriented conservers; Cluster 4 = efficiency-oriented conservers.
Following the clustering process mentioned above, a one-way ANOVA with a post hoc comparison using the Turkey honest significant difference (HSD) test was performed to further profile the clusters according to continuous psychological, economic, and demographic variables affecting electricity conservation. Table 3 summarizes the one-way ANOVA results. The results in Table 3 indicate that there is a significant difference between all four clusters and nearly all continuous variables. The clusters differ in all the sociopsychological variables, except for the ascription of responsibility variable (F = 1.301; p > .05). This indicates that the segment’s consumers might have virtually the same perception about their responsibility toward the electricity problem.
Analysis of Variance Results Comparing Cluster Means Between Continuous Variables.
*Significant at the p < .05 level.
**Significant at the p < .01 level.
Pearson’s χ2 test was carried out to profile the clusters according to the selected sociodemographic characteristics and household features, except for monthly electricity expenditure, which is a continuous variable. Table 4 presents the percentage of each subgroup within the four segments. The results in Table 4 reveal that the segments do not significantly differ in terms of the level of education, gender, type of electric meter used, and the dwelling types (p > .05). Thus, these variables were not used to define the segments. Each cluster was subsequently labeled and described.
Pearson χ2 for Comparing Clusters Between Nominal Variables.
Note. There was therefore no significant difference between clusters within the groups in the variables: type of house, electric meter, gender, and educational level.
*Significant at the p < .05 level.
**Significant at the p < .01 level.
Description of the Identified Four Segments
Cluster 1: Devoted conservers (18%)
This first cluster consists of heads of households highly committed to conserving electricity through curtailment actions and efficient investments. It represents the smallest segment (N = 113). They have a positive attitude toward electricity conservation and respond positively to both the internal pressure to save electricity and the external social norms related to mindful electricity consumption (Figure 1). Even though they already conserve electricity, the devoted conservers are still willing to save more electricity, certainly because they feel guilty for not conserving enough electricity. Surprisingly, with an average of R1,171 (US$83) spent for the electricity bill per month, they are the second highest spenders across the four segments. However, the results show that they have considerably reduced their electricity consumption following the successive electricity price increases.
Graphic representation of segments.
In terms of demographics, most of the devoted conservers are between 36 and 45 years old (29.2%). They are predominantly Whites (50.5%), living in suburbs (78.6%), with a gross monthly income ranging between R25,000 (US$1,785) and R50,000 (US$3,571; 33%).
Cluster 2: Unconcerned nonconservers (28.7%)
These consumers form the second largest segment (as indicated in Table 5) (n = 181). They do not engage in any dimension of electricity conservation and have no intention to make any efforts to use electricity more wisely. The unconcerned nonconservers have a negative perception of saving electricity, because of the constraints and discomforts associated with such actions. They are not under the influence of moral and social pressure to save electricity in their homes and therefore do not feel guilty about not saving electricity (Figure 1).
Comparison of Segments Based on Sociodemographic Characteristics.
Sociodemographically, the majority of these consumers live in suburbs and are tenants. They are predominantly Black, and their age is mainly between 18 and 25 years (31.7%). They are mostly found in the middle class, earning between R10,000 (US$714) and R25,000 (US$1,785) per month (29.5%), and spending on average R922 per month for electricity. Regrettably, they constitute up to 28.8% and appear to be the critical segment toward which to channel electricity conservation campaigns.
Cluster 3: Curtailment-oriented conservers (29.3%)
The curtailment-oriented conservers represent the largest segment (n = 184). They prefer to adopt electricity conservation by curtailment rather than electricity saving by efficiency. Notwithstanding their willingness to conserve more electricity, which is motivated by a positive attitude toward electricity-saving actions, they do not invest in efficient electricity actions even though they are from the middle-income class. Interestingly, they acknowledge being under internal and external normative pressure to conserve electricity, which results in a reasonable amount of guilt when they fail to do so (Figure 1). Curtailment-oriented conservers have slightly reduced their electricity consumption following the electricity price increases. Although they mostly reside in suburbs (58.7%), a considerable number of them live in townships (41.3%) compared to other segments. With approximately R918 (US$65) allocated monthly for the electricity bill, they are the lowest electricity spenders across the segments. They are predominantly Black people (71%), from the middle-income bracket.
Cluster 4: Efficiency-oriented conservers (24%)
The consumers in this segment essentially engage in electricity-saving measures through efficiency improvements, which require some cash investments (Figure 1). This is certainly because they mostly own their houses (83.1%) Sardianou (2007). However, they have no intention to conserve more electricity, and they believe that electricity conservation entails discomfort. Therefore, they neither feel any social norm pressure nor perceive any moral pressure to save electricity.
Of all the segments, efficiency-oriented consumers have the highest monthly electricity bill (R1,315 [US$94] on average). In fact, they are likely to take no action even after consecutive electricity price escalations. They are mostly rich people (nearly 61% earn more than R25,000 [US$1,785] per month), living in suburbs (70.2%). Therefore, their investment in electricity-saving actions by efficiency seems to be motivated by reasons other than the desire to reduce the electricity bill and save money. While this segment mostly consists of Black people (53%), there is also a substantial percentage of Whites (33.1%) and a nonnegligible proportion of coloreds (7.9%) compared to other segments. The characteristics of the four identified segments are graphically presented in Figure 1.
Discussion
Considering the electricity problem in South Africa and the general lack of empirical research from developing countries on energy conservation and consumer behavior, this study examined electricity conservation segments in South Africa. The study used behavioral bases of segmentation with the purpose of guiding social marketing campaigns aimed at reducing electricity consumption in households. The results of the cluster analysis revealed four segments. The four segments were subsequently labeled as devoted conservers (18%), unconcerned nonconservers (28.7%), curtailment-oriented conservers (29.3%), and efficiency-oriented conservers (24%).
It is important to note that some of the segments found in this study are quite similar to those found in prior energy conserver segmentation studies carried out in other settings. For example, the devoted conservers in this study are similar to the “idealistic energy-savers” found in Switzerland by Sütterlin et al. (2011). This segment engages the most in electricity-saving efforts based on both curtailment and efficiency actions. Consumers in this group are willing to save more energy and have a positive attitude toward electricity-saving actions.
The findings in this study have theoretical and practical implications for the understanding of consumer conservation behavior. The study contributes to addressing the theoretical questions in literature about curtailment and efficiency as dimensions of energy conservation behavior (Karlin et al., 2014). The findings are in line with arguments in studies by Karlin et al. (2014) and Botetzagias, Malesios, and Poulou (2014), who have noted the distinctiveness of dimensions of conservation behaviors. The results particularly indicate that in the case of electricity conservation, while some electricity users are simply nonconservers and others are both curtailment and efficiency oriented, most consumers are either curtailment or efficiency oriented. The findings show that in the South African context, not only are the two dimensions distinct but they also associate, respectively, with a significant number of consumers. Moreover, this finding points to the fact there are barriers (e.g., insufficient income, lack of knowledge of energy efficiency products) that deter curtailment-oriented conservers and some of the unconcerned nonconservers from investing in energy efficiency. In the same way, adopting electricity curtailment actions entails some discomfort that might hamper efficiency-oriented conservers and some of the unconcerned nonconservers in their attempt to conserve electricity.
In narrowing the notion of spillover within the two dimensions of electricity conservation to some extent, this finding indicates that for the majority of electricity consumers, the adoption of one dimension of electricity conservation does not necessarily implicate the other. The segments curtailment conservers (29.3%) and efficiency-oriented consumers (24%), which represent about 53.3% of our sample studied, were involved in only one of the two dimensions of electricity saving. This opens a debate about the significance of the spillover effect within the dimensions of electricity conservation.
From a practical perspective, the findings in this study can be used to guide the formulation of effective social marketing programs aimed at promoting electricity conservation behaviors. As in any marketing effort, social marketers need to have a clear understanding of the target market. More attention needs to be granted to the targets “most worth pursuing” (Yankelovich & Meer, 2006, p. 2). Weinreich (2010) describes two categories of target that deserve more consideration in social marketing. These are the target of risk and the target of opportunity. Targets of risk in this case are segments that are less likely to engage in electricity-saving actions. The two segments unconcerned nonconservers and to some extent efficiency-oriented consumers can be considered targets of risk, because the former do not engage in electricity-saving actions and have shown no intention to conserve electricity, while the latter have a high living standard and therefore appear unmoved by the electricity price increases, having the highest electricity bills (and consequently the highest electricity consumption). These efficiency-oriented consumers seem careless in consuming electricity despite investing freely in electricity-saving equipment. With the satisfaction that they have played their part in buying an electricity-saving device, they develop antipathy for the whole concept of electricity conservation, as evidenced by their negative attitude toward it. This behavioral response is known as the “rebound effect” and suggests that one-time energy efficiency improvements may fall short due to subsequent behavior changes (Sorrell, 2007). Studies in the UK (Sorrell, 2007) and South Africa (Davis, Cohen, Hughes, Durbach, & Nyatsanza, 2010) note that the rebound effect should be factored into the design of energy efficiency initiatives.
Reaching out to these two targets of risk would therefore entail finding effective ways of changing their attitude toward electricity conservation. This could be done by showing additional benefits related to electricity conservation, such as environmental benefits (Dogan, Bolderdijk, & Steg, 2014). The positive influence that consumer perception of benefits associated with behavior has on attitude is widely acknowledged in literature (Ha & Janda, 2012; Lanzini & Thøgersen, 2014; Rodríguez-Barreiro et al., 2013). In looking at benefits, it is important to note that some benefits of conservation behavior accrue to society at large, while others have a direct effect on the individual. In promoting the benefits, social marketers need to ensure that their communication strategy touches on both types of benefit so as to maximize potential impact. Unconcerned nonconservers and efficiency-oriented consumers need to be reminded that it is not only in the interest of society but also in their own interest to conserve electricity. Ohler and Billger (2014) found that personal self-interest (e.g., financial gain) has a much more significant impact on energy-saving behaviors than social interests. Financial incentives or disincentives can be helpful in reducing the electricity demand in these segments. Prior studies have proven, for example, the dissuasive power of financial sanctions on heavy electricity consumers (see Senbel, Fergusson, & Stevens, 2013). Dissuasive pricing strategies such as the IBT should therefore be supported.
Apart from benefits, social norms have been proved to foster electricity conservation effectively (Grønhøj & Thøgersen, 2011; Kleinschafer & Morrison, 2014). Allcott (2011) found that knowledge of a neighbor’s low electricity consumption relative to one’s own helps one to think about reducing electricity consumption. Social marketers could thus activate social norms and moral pressure as a way of triggering willingness to conserve electricity. Norms can be activated by demonstrating that most people (friends, neighbors, celebrities, and community members) use electricity sparingly in their home.
“Targets of opportunity” refers to those segments that are ready and willing to engage in electricity-saving measures (Weinreich, 2010). The curtailment-oriented conservers segment falls into this category, consisting of people who already make some effort to conserve electricity through curtailment actions and are willing to save more electricity but have failed to invest in more electricity efficient actions. Fortunately, the curtailment-oriented conservers represent the largest segment of electricity conservers in South Africa and are the lowest electricity spenders. Given the significant size, responsiveness, and accessibility of this segment, the curtailment-oriented conservers should be an important target of interest in terms of supporting and fostering their conservation behavior. This could be done by providing financial rewards for conserving electricity by curtailment actions. Nonfinancial rewards could include public acknowledgment of electricity conservation efforts in the community.
In addition to encouraging curtailment-oriented conservers to sustain their conserving attitude and behavior, they should be encouraged to incorporate efficiency-oriented actions despite their remarkable commitment to curtailment actions. This could be done by finding ways to remove or reduce barriers (especially financial barriers) to their undertaking of such activities. Policy makers could, for example, facilitate access to low-interest loans for investing in technical adjustments and energy efficient appliances to conserve more electricity at home. They could incentivize investments in efficiency products, such as solar geysers or gas heaters installed for efficient consumption.
While it may be relieving that up to 18% of the samples are devoted conservers, it is worrying that their electricity bill per month was the second highest. The high bills could result from the fact that they own many electrical appliances and equipment, since they are fairly wealthy, or that they do not monitor how much they are spending even though they make efforts to conserve. The solution here could be the provision of regular feedback on their electricity consumption and its possible impact (Grønhøj & Thøgersen, 2011). Barriers hindering the devoted conservers from enjoying the benefits of their electricity curtailment and efficiency efforts need to be identified and dealt with.
Conclusion, Limitations, and Suggestions for Future Research
From this study’s findings, it can be concluded that using behavioral segmentation to divide electricity conservation behaviors into curtailment and efficiency segments reveals the types and levels of effort that South African consumers are making to conserve electricity. Profiling the segments according to sociodemographic characteristics and certain economic and psychological drivers provides a richer picture of the types of people making or not making conservation efforts. While the findings will be of theoretical and practical importance to social marketing scholars and energy-conserving policy makers, the study is not without limitations. The first limitation relates to the use of self-reported measurements to assess the extent to which participants conserve electricity. This leaves room for social desirability bias when completing the questionnaires (Kreuter, Presser, & Tourangeau, 2008). The second limitation is that electricity conservation by curtailment and by efficient investment presents a descriptive frame for understanding the barriers and benefits of adopting electricity conservation. Further study should focus on this level of granularity. Another limitation is that this study focused essentially on the big cities in the Gauteng province. To overcome these limitations, future research could cover other urban areas and add some rural areas. Doing so would help increase the external validity of the findings. Moreover, future studies could make use of experimental research design so as to have an unbiased and more realistic picture of the electricity-saving activities.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflict of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the Global Excellence and Stature (GES) through the University of Johannesburg.