Or if you're not pay- ing their bill, then you don't need to know. .... issues], they don't want to be viewed ... Vaughan, T. The landlord-tenant relation in a low-.
In the Dark, Out in the Cold For 30% of the population, lack of access to home-energy monitoring devices translates into a lack of power—in more ways than one. By Tawanna Dillahunt and Jennifer Mankoff
S
DOI: 10.1145/1961678.1961685
ocioeconomic factors play an important but largely hidden role in home-energy consumption. Most studies have targeted single-family, affluent households—so the resulting energy-monitoring systems do not address the needs of renters and low-income individuals. For technology designers and information architects, failure to address the disparity in the use of home-energy devices can no longer be set aside. Lack of access to energy information leaves three out of 10 U.S. households without the power to make changes in consumption. To further explore and identify socioeconomic factors affecting energy usage, we conducted two studies. In the first, we looked at energy use in low-income households; in the second, we focused on the impact of landlord/tenant relationships on energy use. Our population included a broad range of household types that included one person, a family or extended family, roommates and long-term visitors, and ownership relationships that included landlords/renters, governmentrun buildings, and homeowners. Our results demonstrate the importance of socioeconomic context: A closer look at the dynamics of lowincome households reveals that beyond the heads of households living in a home, energy use may be affected by neighbors, other members within a community, and landlords. All of these stakeholders affect energy consumption directly and indirectly. In addition, an individual’s beliefs, cultures, household structure, and the availability of money may all affect energy consumption. Landlords may pay for energy consumption, and low-income individuals may receive stipends from the government. In some cases, a
XRDS • SUMMER 2 01 1 • V OL .17 • NO.4
household will be allocated a certain number of kilowatt-hours per month. The complex effects of these factors require a deeper look at what is going on in low-income and rental households. In this article, we provide an overview of what we learned in our two studies and discuss how this information could influence the design of energy feedback technologies.
STUDY 1: ENERGY USE IN LOW-INCOME COMMUNITIES Heating, cooling, and electricity use in the home is one of the biggest forms of personal energy consumption, especially for low-income households. To help us understand home-energy consumption in these households, we recruited 26 low-income residents across a small rural town in eastern North Carolina and a large metropolitan area in Pennsylvania [2]. We recruited tenants from government-subsidized apartments and low-income neighborhoods. We gave participants disposable cameras and asked them to take pictures of anything reminding them of energy consumption. Participants used their own definitions for “energy.”
During our interviews, we found that participants often did not pay their own utility bills (or only paid a portion of them). We were also able to identify multiple stakeholders that play a role in household energy consumption and explore participants’ motivations for saving energy and identified barriers they encountered. We discuss our findings here. Alternate billing methods. In rented apartments or very low-income households, individuals are not always responsible for paying their own energy bills. For example, in some cases, landlords or public-housing authorities factor energy bills into the rent. In other cases, low-income individuals and/or tenants receive a bill only if they consume more than an allocated amount. Others may receive monthly stipends to aid them with monthly energy bills, while some may divide or share the total cost of the electricity based on the number of apartments on a particular floor. These myriad models suggest future natural experiments in the impact of different types of financial motivation on energy use. In many of cases, the building infrastructure determines the cost model. For example, if a building is mastermetered, there is no way to designate how much an individual household consumes. This implies a real need for sensing devices that are capable of gathering energy use information across master-metered units. Such cost structures encourage landlords as well as some tenants to save energy. Perhaps, surprisingly, many of the lowincome households we interviewed saved energy even though they were not responsible for payment. Why would residents save energy even if overuse cannot be traced to them? Motivations for saving energy. Financial concerns did not always drive individual motivations. Of course, not all participants were interested in saving energy, and some felt they were “doing enough already.” However, many participants saved energy for reasons spanning comfort, the desire to protect the environment, and spirituality. For example, some participants cited their connection to God as a key reason for avoiding waste and saving energy. 39
Other tenants’ concerns for the environment came from their responsibility to future generations. For instance, it was important for one tenant to save energy “[for our daughter], and her children and grand children.” In fact, many participants’ concerns for future generations extended past the environment. For example, one participant volunteered for Head Start and a sustainable garden program, in addition to helping to run a local crime prevention event. Current home-energy monitoring systems display information about how much energy is being used and what the projected monthly bill will be, that is, the costs of energy consumption [3]. However, there may be opportunities for devices to suggest more ways to save. Additional motivational factors such as saving future generations could be especially helpful for individuals who feel they are doing enough already. Multiple stakeholders. Beyond the heads of households, landlords and other members of the community affect energy consumption. In fact, children, friends, guests, and other household members are often barriers to saving energy. Concerns about safety and theft affected how participants used lights; a community could kill or encourage a neighborhood garden. Energy savings requiring infrastructure investments depend on landlord buy-in. Thus, energy use and feedback must situate itself within the context of a community of people who affect energy use. As a result, communication among multiple stakeholders could be a beneficial feature to add to today’s energy monitoring devices. Individuals from our smaller, rural areas, for example, shared information about energy bills and energy-saving strategies with their community. As one participant stated, “…my cousin, I can tell you her bill was $400 and some change. My friend down there, her bill was $300 and some change. My aunt’s light bill was three and some change. I guess it’s just because we could, because you know what I’m saying, we compare stuff…” Those from the more metropolitan area did not discuss their consumption or behaviors with others for fear of being too intrusive. One resident stat40
A scenario describing how an individual may use a system designed for sharing energy consumption.
ed, “I guess ’cause people think you are in their business. Or if you’re not paying their bill, then you don’t need to know.” Nevertheless, individuals still desired this information. Energy feedback is a multi-user system, and this realization leads to many design opportunities. We could design a system to explicitly engage children, or adjust the heat based on the preferences of multiple household members and on who is present. A feedback system could enable the sort of communication described by the participant above (sharing information about energy costs, discussing possible reasons, and so on). One possibility that has already received some study is to explicitly encourage participants to compare energy use with each other. In the scenario shown in the figure above, the character can see how his neighborhood compares with others, drill down to see how his consumption compares to others in his apartment building, and discuss his consumption with members of his community. Comparisons between individuals or groups when combined with feedback about performance can be especially useful [4]; however, there are varied thoughts on how effective social comparisons in environmental psychology will be [3]. For example, although comparisons generate interest and engagement, they do not always lead to behavioral changes. Barriers to saving energy included structural inefficiencies, from building infrastructures, to older drafty windows and doors, to poor insulation. Another key barrier was financial constraints. For low-income homeowners, tax-break incentives requiring upfront capital investments in infrastructure improvements are ineffective. For
renters, the landlords are responsible for making upfront investments such as building infrastructure and efficient appliances. Finally, participants wanted more information about how much energy they used, but those with subsidized energy or those in mastermetered apartments did not receive a bill or feedback of any kind. Introducing these factors into energy-monitoring technologies may create new problems, some of which we explore in our second study. In particular, since energy savings depend on infrastructure investments in addition to day-to-day behaviors, there are important implications for the landlord’s role in energy savings.
STUDY 2: LANDLORD/TENANT CONFLICTS AROUND ENERGY While it may be straightforward to engage stakeholders within a household in energy consumption conversations, the role of those outside of a household is murkier. In our second study, we revisited the data about landlord/tenant relations from our first study, and contextualized it with data from tenant focus groups and landlord interviews [1]. Following the same methods as before, we recruited landlords from a website listing landlords who rented to tenants receiving public assistance. Power imbalances. Tenants are relatively powerless in the landlord/ tenant relationship. To understand the tenant’s perspective, we need to understand that tenants—particularly low-income tenants—may feel lucky just having a place to live [5]. As a result, some tenants fail to report their needs to their landlord because of factors such as income and status. One resident stated, “I think I have a hole in the wall and that’s where air is XRDS • SUMMER 2 01 1 • V OL .17 • NO.4
coming from but they’re not going to do anything about it…It’s public housing. If it was a house, they sure will find anywhere the air is coming from.” Our study showed that some residents who were more knowledgeable about what could be expected were able to advocate for changes, particularly when they understood tenant rights and how to negotiate with their landlords. Landlords also felt a level of powerlessness in the relationship. Most landlords we spoke with felt that tenants at times took advantage of them. Landlords especially felt taken advantage of when they saw resources wasted that tenants did not pay for. For example, one of our landlords described a situation where a tenant waited to notify him about a broken thermostat and instead opened his windows in the winter because he was too hot. This delay in notification meant paying for extra heat (until the tenant reported the issue, and then until landlord found someone to address the issue). Many of the landlords we interSummary of results and design implications. Factors affecting energy use design implications
Design Implications
Sensing devices should Alternate billing types add difficulty to be capable of gathering monitoring energy use. energy use information across master-metered units. Motivations for saving energy vary.
Multiple stakeholders impact home energy consumption.
Integrating motivations into energy-monitor displays (for example, money, spirituality, the desire to protect the environment and future generations) could help designers expand technologies for new audiences. Displays that allow individuals to compare consumption, hold discussions around consumption, and that suggest actions related to reduced energy use could be beneficial in some neighborhoods.
Broader audiences face Consideration of barriers that prevent new problems as a energy conservation. result of barriers to saving energy (for example. financial constraints, building infrastructures, lack of tax-incentives) may have a positive impact on broader audiences. Power imbalances Consideration of the prevent disclosure of power imbalances problems and solutions. between stakeholders such as landlords and tenants could lead to greater mutual understanding of how energy is used (and wasted).
XRDS • SUMMER 2 01 1 • V OL .17 • NO.4
viewed tried to explain this perceived wastefulness. Landlords explained that “[residents] don’t care because they are not paying utilities…there’s no way to force them to be energy efficient when they don’t care,” that “people feel uncomfortable [with regard to raising issues], they don’t want to be viewed as a complainer.” Or, “maybe it’s their upbringing.” These explanations demonstrate a range of assumptions about tenants (lack of caring, fear of retribution, and resignation) that contrast with what residents reported. One approach to resolving these differences is to share information across stakeholders. For example, one study participant described how high building wide heating bills led to a group of tenants advocating for building improvements. Feedback technologies could help enable community action by making it easier to identify common issues (see the figure), or improve landlord tenant communication about shared resources. Many of the issues found in our second study come down to differences not only in financial responsibility, but also in the availability of information. For example, the advocacy just described depended on tenants knowing that they all had high heating bills. Similarly, giving landlords better information about each apartment’s energy use could address their dissatisfaction with tenant wastefulness. While the positives of sharing information in this way are clear, we must also consider the negatives. Information can affect the balance of power between parties [1], and this may lead to negative outcomes. For example, allowing a landlord to view each apartment’s energy use may be a breach of privacy that could lead to censure or unexpected disclosures. Similarly, sharing between apartments might affect a person’s sense of security (for example, we might be able to guess when she comes home each day). Finding a way to support communication without creating negative outcomes is a challenge for future designers of energy feedback systems.
CONCLUSION Designers should consider socio-economic factors, alternate billing methods, and multiple stakeholders when creating eco-visualizations to reach
new audiences. We summarize our design recommendations and their relationship to our studies in the accompanying table. Energy use and energy-saving behaviors take place across all sectors of our society. Any energy-saving solution that only considers the needs of less than half the population, (that is, affluent households) or fails to address the true complexity of living situations, (for example, the presence of landlords and other household members) is limited in its potential impact. Moving forward, researchers should make explicit the socio-economic group being targeted/ studied. Projects that explicitly broaden their socioeconomic reach have the potential for a larger impact (in terms of the number of people they can reach). Studies of low-income communities may also help inform the design of technology for affluent households by identifying new motivations, uncovering ways of coordinating with other stakeholders, and identifying new financial models that may affect behavior. Biographies Tawanna Dillahunt is a Ph.D. student in the Human Computer Interaction Institute at Carnegie Mellon University. Her research focuses on improving the behavior of individuals using ubiquitous and social technologies particularly in the domain of environmental sustainability. Jennifer Mankoff is an associate professor in the Human Computer Interaction Institute at Carnegie Mellon University. She received her Ph.D. in computer science from the Georgia Institute of Technology, where her research focused on the role of interactive technologies in addressing critical social problems. Her work investigates reusable tools and techniques that apply to mobile, desktop and social web technologies. References 1. Dillahunt, T., Mankoff, J. and Paulos, E. Understanding conflict between landlords and tenants: Implications for energy sensing and feedback. In Proceedings of Ubicomp ’10 (2010), 149-158. 2. Dillahunt, T., Mankoff, J., Paulos, E. and Fussell, S. It’s not all about green: Energy use in low-income communities. In Proceedings of Ubicomp ’09 (2009), 255-264. 3. Egan, C. How customers interpret and use comparative displays of their home energy use. In Proceedings of European Council for an EnergyEfficient Economy (1999), Panel III, 19. 4. Froehlich, J. Findlater, L., and Landay, J. The design of eco-feedback technology. In Proceedings of CHI ‘10 (2010), 1042-1052. 5. Vaughan, T. The landlord-tenant relation in a lowincome area. Social Problems 16 , 2 (1968), 208-218. 6. Weber, C. and Matthews, H.S Quantifying the global and distributional aspects of American household carbon footprint. Ecol. Econ 66 , 2-3 (2007), 379-391.
© 2011 ACM 1528-4972/11/0600 $10.00
41