Building Student Sustainability Competencies through a Trash-Practice Nudge Project: Service Learning Case Study in Kuwait

Abstract

This discussion describes an experimental behavioral economics class implemented in a service learning format. Students implemented two nudge interventions to influence public trash behavior, which is an issue throughout the Middle East/North Africa (MENA). The aim in one project was to encourage more use of trash bins in a multi-screen theater and in the other to sort plastic, glass, and paper when throwing trash in the public bins at a university. These two organizations paid the implementation costs, which were quite low, as common for most nudges. The class was co-taught by two university faculty and several personnel from governmental and UN offices responsible for building nudge capabilities in Kuwait. In each case, results in the student projects demonstrated that nudging resulted in a substantial improvement. Students gained strong competencies in applying sustainability principles to a specific problem and in carrying out a real-world project. They learned the importance of collaborating with stakeholders and got to see that their work was actually used in policy formation by the Kuwait government and the UNDP’s Kuwait office.

1. Introduction

This discussion examines the implementation of an experimental behavioral economics class conducted at a university in Kuwait, which aimed to help build student competencies in understanding sustainability issues and implementing small sustainability projects. The class was built around two student-designed research projects, which investigated nudge interventions to influence consumer public trash behavior by encouraging more beneficial subconscious choices for behaviors that are rarely carefully considered [1]. One aimed at increasing the use of indoor trash bins to reduce cleanup costs from littering, and the other aimed to encourage the sorting of materials at outdoor public trash bins. Nudging is often effective at gaining small changes in habitual behaviors that people may not consciously think about (e.g., [1]), including public trash behavior (e.g., [2]). The nudge impact is described here, but much of the discussion aims somewhat more broadly at covering several intersecting themes, which can come together to help deal with serious waste management problems common in most developing economies:

• influencing consumer trash behaviors in public places;
• action learning to give students practical experience in research about real-world problems;
• developing sustainability mindset and competencies among students who will soon graduate;
• university and community cooperation to involve stakeholders in addressing such problems.

In general, the nature of key environmental problems is similar throughout Gulf Cooperation Council (GCC) countries [3], and waste management aspects are no exception [4]. Zafar ([5], pp. 162–163) talks about an “obsession with landfills” and notes that very little recyclable plastic, paper, or metal is actually recycled, with most ending up in landfills. A relative lack of recycling infrastructure is one key reason for this, but Zafar [5] also notes that a lack of source separation hinders recycling, making it somewhat less economically attractive to invest heavily in recycling. This is a worldwide issue [6], because poor consumer cooperation in sorting increases the costs of waste management (e.g., [7]). The student research projects discussed here specifically assessed nudge methods to encourage greater cooperation in consumer public trash disposal behavior, including sorting in one project.

Nudging basically blends methodologies from behavioral sciences, especially economics, psychology, marketing, to understand human behavior and then develop nudge policies, which can encourage beneficial behavior by consumers through addressing their System 1 subconscious level thinking. Such methods are useful because the lack of conscious-level System 2 thinking is very common for routine actions in daily life. Nudges can influence behavior without consumers necessarily having to think about them carefully.

“A nudge … is any aspect of the choice architecture that alters people’s behavior in a predictable way without forbidding any options or significantly changing their economic incentives. To count as a mere nudge, the intervention must be easy and cheap to avoid. Nudges are not mandates. Putting fruit at eye level counts as a nudge. Banning junk food does not”.

([1], p. 6)

The projects were implemented with active involvement by the clients and governmental offices involved in nudge work in Kuwait. There is growing consensus that education about sustainability is most effective when it incorporates some degree of action learning and practical experience (e.g., [8]). Claro and Esteves [9] argue that active learning approaches build “sustainability-oriented capabilities” among students, which they categorize into three dimensions: content and methodological expertise; practical knowledge that can be implemented; and ability to effectively collaborate with different stakeholders. Service learning is one form of active learning, incorporating projects beneficial to the community to facilitate an understanding of course content and enhance civic engagement [10]. Service learning may involve projects in the private and public sectors and a broad range of stakeholders.

The student projects here were implemented in collaboration with the Kuwait Policy Appraisal Lab (KPAL, Kuwait Government), the Qatar Behavioural Insights Unit (QBIU, Qatar Government), and the United Nations Development Programme (UNDP) in Kuwait. Their participation was part of a regional effort led by the QBIU head to develop nudge capabilities throughout MENA [11]. Students also had to work closely with the local clients for the two interventions, which were a cinema and the university.

This experimental course was part of a broader long-term initiative to integrate more active and service learning into university classes. To date, a number of studies have found that such practices are associated with variety of benefits, including enhanced civic engagement and social responsibility among students, and useful small projects that help community organizations improve operations [10,12]. The projects here involve training students in the practical implementation of research projects about key resource management issues. Aljamal et al. [13] discuss this initiative for water resources, but the broader program includes all three key resource issues in MENA: water–energy–food, often called the WEF nexus [14,15]. All three are closely connected to environmental issues and require coordination with a range of stakeholders. Notably, waste management is a critical issue for all three, even to the extent that some observers talk about the nexus as consisting of four elements: water, energy, food, and waste management [16]. Essentially, for one little piece of this huge problem, the course aimed to go well beyond just knowledge of these particular sustainability issues or even toward developing a sustainability mindset and competence in dealing with sustainability problems [17,18].

The context of this discussion is Kuwait, essentially a city-state, at the top of the Arabian Gulf. It has been an important port city for several centuries (e.g., [19]), although nowadays, urban life has been strongly conditioned by oil wealth [20]. Some observers see a strong connection between affluence and trash production (e.g., [21,22]), and Kuwait certainly seems to fit this pattern. “High standards of living and rapid economic growth has been a major factor behind very high per capita waste generation of 1.4 to 1.5 kg per day” [23]. The country was already facing problems with solid waste landfills a quarter century ago [24]. Now, its many landfills are mostly poorly managed, running out of capacity to take more trash, and causing public health and environmental problems [5,23]. Such problems are not unique; rather, they are common in MENA countries [5], as well as in many developing economies even outside the MENA.

2. Literature Review

This section briefly examines conceptual issues behind the four themes noted in the introductory paragraph. Consumer trash behavior is the real-world problem examined in this discussion (and the class). Better trash sorting, of course, will not entirely solve the huge waste management problem in MENA [4,5]. It does, however, reduce the cost of recycling [7]. Low-cost nudges of the sort discussed here rarely entirely fix things, but they very often reduce problems and contribute to better environmental practices. Beyond that, service learning approaches in classes give students practical experience in dealing with real-world problems. A sustainability-oriented service learning project helps develop a sustainability mindset and competencies among students [17,18]. In the process, they participate in university–community cooperation and learn the importance of involving stakeholders in projects that address environmental issues [25,26].

2.1. Consumer Trash Behavior in Public Places

The specific projects students implemented were about nudging consumers to use trash bins more consistently and (on one project) to sort when doing that. Waste management has become a serious environmental issue in MENA [4]. There is not very much recycling of plastic, paper, or metal, at least from the consumer side of using these materials. As noted in the introduction, some observers lament an “obsession with landfills” in MENA ([5], pp. 162–163). The need for better sorting in MENA is occasionally examined, often for materials that can be used in composting [27,28]. To date, however, the somewhat sparse literature on the issue is largely about environmental impacts or technical issues of waste management systems [29].

There is only scattered research anywhere in the world specifically about consumer sorting of (mainly) packaging materials waste. Much of what there is focuses either on conscious thinking by consumers (e.g., [30,31]) or technical solutions related to the packaging itself (e.g., [32,33]). In general, however, a comprehensive view must necessarily treat urban waste management as a wicked problem (e.g., [34,35]). Consumer behavior regarding packaging is a very specific example of one such wicked problem [36]. The nature of wicked problems has been recognized for quite a while [37]. They are more-or-less intractable, highly complex, and thoroughly interconnected with a whole myriad of factors [38]. Part of this wickedness is simply that it is often difficult to get people to think consciously about small habitual actions [1]. Discarding trash after a snack, for example, may not be an action that rises above subconscious habit to careful, conscious thinking. Better infrastructure, better packaging, more education, etc., can all help somewhat, but they cannot fully solve this problem

Jesson [39], for example, pointed out that there are different levels of commitment to recycling and various barriers. Situational barriers (such as a lack of facilities for sorting) can reduce beneficial behaviors even among strongly committed people. A lack of knowledge and awareness is another problem, as are various attitudes about whose responsibility recycling is—individual consumers, companies, government, etc. Jesson ([39], p. 26) points out that “recycling marketing has been predominantly knowledge- and education-based”. Often it is one-way, with little attention to how average consumers actually think about the issues—if they actually consciously think about them at all. AFED [40] indicates that Kuwait actually has fairly good environmental education, but the report does not say much about recycling specifically.

Good education, however, will not necessarily raise subconscious habitual behavior to conscious thinking about the environment. Aljamal et al. [41,42] confirm that Kuwaiti citizens are knowledgeable about the environmental impacts of wasteful water use, to cite a specific environmental issue. But water usage is routine habitual behavior. Knowledge of environmental impacts is not top-of-mind when confronted with suggestions that Kuwait’s nearly full water subsidies should be reduced to discourage wasteful use. The top-of-mind response to paying for water is strongly negative. With more depth of discussion, however, most people feel that paying more for water makes environmental and financial sense and is acceptable.

Jesson’s [39] behavioral barriers mostly are essentially what might be called routine, System 1 thinking in nudge terminology [1] or low involvement thinking in marketing language [43,44]. Thinking about recycling does not rise above the subconscious level for many people, so they do not recycle, even though they might if they did actually think about it. Jesson ([39], p. 35) notes the “difficulties in establishing routines for sorting waste”. Van Doorn et al. [45] discussed aspects of this important issue some time ago. Weak attitudes, even if favorable, do not translate into behavior very well, while strong attitudes usually do. A recent review shows that this non-linear relationship is fairly common [46]. It is easy to see that nudge techniques might help with our problem, trash sorting. Strongly pro-recycling people are likely to sort without any prompting. Weak-attitude people probably will not. They may not be against sorting at all, but it simply does not come to top-of-mind conscious-level thinking in most situations where they might be able to sort.

In other words, most people probably do not consciously decide that they want to harm the environment by, for example, littering or failing to sort to make recycling easier. In Jordan, for example, Moqbel et al. [47] indicate that people cite a number of strong reasons not to litter (but also are not very concerned about several other reasons). Nevertheless, this often does not translate into actual behavior. Qualitative research in Saudi Arabia shows similar results [48]. Many issues relate to Jesson’s [39] barriers. Some reasons simply reflect the difficulty of actually shifting away from habitual behaviors when something is only at the subconscious level rather than top-of-mind. There is no easy solution, in Kuwait or more widely—consumer public trash behavior is more-or-less habitual.

Building on prior research, Jesson [39] proposed a simple model, which we adapt here (Figure 1), to illustrate the problem. The figure shows levels of awareness among consumers about recycling issues and awareness of what consumers themselves could do about them. At level 1, if consumers are completely unaware of issues or actions, education can help in starting to move them.

Relatively small things, however, do not get a great deal of conscious attention, and education may not move thinking into the strong attitude range, which Van Doorn et al. [45] demonstrate is necessary to translate attitudes into action. Almosa et al. [48] and Moqbel et al. [47], for example, show that intrinsic motivations are stronger than extrinsic motivations for getting cooperation in using litter bins. Education—informing consumers why they should do things—is essentially extrinsic. Consumers may pay attention to things that they consider quite important or to things that temporarily catch their interest. It is less likely to work for small things they rarely think about.

And, even if education does succeed in making people consciously aware of the need for sorting/recycling, at level 2 they may still not know exactly what they can do to cooperate. Even more education can help, but that is a relatively slow and somewhat expensive solution, which does not overcome the subconscious nature of most everyday routine actions. Nudging can start gaining cooperation quickly and cheaply. The research projects discussed here show this—they specifically assessed nudge methods to encourage greater cooperation in consumer public trash disposal behavior.

2.2. Service Learning to Develop Nudging Competencies

The experimental service learning class discussed here was built around these projects. As noted above, service learning is a subset of experiential learning, focusing on enhancing students’ theoretical knowledge through application in a real-world project that serves the community [10]. Service learning did not get much academic attention outside the US until the last decade or so [49], but in the US, publications started appearing more frequently in the 2000s. Ziegert and McGoldrick [50], for example, regard basic economic competency as a life skill and argued that service learning promoted deep rather than surface learning, making it more likely that students would apply useful economic concepts in their lives. Sipos et al. [51] applied the ideas to sustainability, criticizing the common mostly theoretical and classroom-based educational methods and arguing that more was needed to make much progress in fostering sustainable behaviors. They proposed learning based on action research to “integrate transdisciplinary study (head); practical skill sharing and development (hands); and translation of passion and values into behaviour (heart)” ([51], p. 68). Hirscher et al. [52] recently applied this simple “head, hand, heart” schema for teaching sustainable consumption skills: learning the theoretical concepts, implementing them in real-world experiments, and reflecting on what the experience shows. In general, this is a version of deep learning, which is often cited as necessary for sustainable development (e.g., [53]).

Service learning goes even beyond this experiential learning approach, incorporating civic engagement, which is sometimes mentioned in experiential learning/action research discussions, but not usually stressed much. Salam et al. [10] do stress civic engagement as part of service learning. They distinguish service learning as intermediate between engagement with the community through volunteerism, which can often give somewhat unstructured real-world experience, and internship, much more structured, formal real-world experience. The implication, more or less, is that too unstructured or too structured can both hinder student ability to learn either the conceptual issues or the competencies quite as well. They point out the benefits from deeper learning about how theory connects to practice, and how community partners in the service learning projects gain some help in dealing with their real problems. Summarizing current literature, Hallinger and Narong [49], say that “service learning incorporates academic study of interdisciplinary content knowledge, community-based activities that provide practical experience, and civic engagement”. This goes beyond just “heart” and self-reflection on what was learned and requires interaction with the community to see what was the benefit in the view of those stakeholders. Such soft skills are now widely recognized as an important aspect of sustainability competency.

Venn et al. [54] categorize key sustainability competencies into two groups: with “sustainability research” constituting the ability to understand and analyze sustainability issues. Basically, these are hard skills, which universities are pretty good at instilling. Sustainability certainly needs a wide range of technical and managerial skills, details of which are somewhat dependent on what specific aspect of sustainability people work in [54]. Most sustainability problems cannot easily fit into specific academic silos, and critical thinking competency is the ability to translate the hard skills into specific contexts [55].

Most actual implementation requires a range of “sustainability intervention competencies” [54]. Sujová et al. [56], discussing sustainable education, cite an employer survey indicating that only 40 percent of big employers in the USA that feel university graduates have sufficient soft skills. “The lack of good communication, collaboration, and conflict resolution skills are holding people back” ([56], p. 4). Some of these soft skills are the “social competencies that enable a manager to interact with diverse stakeholders” ([9], p. 1249). Managers who implement sustainability projects must be able to communicate well and engage stakeholders, notably the organizations helping implement projects, and the people who use the organizations’ products or services (e.g., [57]).

The ability to communicate and collaborate well with stakeholders was considered among the most important competencies in all the prior research. Venn et al. [54] reviewed this, and their own research among sustainability professionals confirmed this. “Interpersonal collaboration lies at the heart of the intervention competencies” ([54], pp. 15). Service learning is excellent at developing the problem-solving skills and collaborative abilities needed to interact with stakeholders [58]. Hebert and Hauf [59] provide strong statistical evidence that service learning contributes well to soft skills competency development, such as civic responsibility and interpersonal skills.

As noted in the introduction, the specific experimental course discussed here was part of a longer-term initiative to integrate student learning into societal sustainability issues [13]. The goal, as in a number of courses in this specific initiative, was for students to understand and be able to analyze problems (“head” in the schema of Sipos et al. [51], to give them the skills set needed to implement little practical projects (“hand”) and to develop enthusiasm for some activism (“heart”) based on experience that they actually can achieve small changes). But students were not the only stakeholders. Aljamal et al. [13] also note the interaction with governmental and quasi-governmental organizations, as well as with local private companies. They all gain small benefits from such collaboration, and students, the future leaders who will carry on work toward sustainability, learn the importance of involving a wide range of stakeholders.

Hallinger and Nguyen [60] identify three main themes in education for sustainable development. Teaching and learning for sustainability are closely related to the main topic in this discussion. They note that some researchers working on these issues aim at action competence, i.e., not just changing students’ behavior on an individual level, but also enabling them to contribute to changing wider society toward better sustainability practice. The second theme, developing a sustainability mindset, which they note is tied in with experiential learning, is also very obvious in these student projects. The biggest (and older) theme is general education for sustainable development, more about basic principles, which the other two research streams focused on implementing. The student projects include elements from all three of these themes—and such interconnection seems entirely consistent with sustainability concepts and building sustainability competencies. The United Nations points out that “the 17 SDGs are integrated–that is, they recognize that action in one area will affect outcomes in others, and that development must balance social, economic and environmental sustainability” [61], and Venn et al. [54] discuss the range of practical skills needed in implementing sustainability interventions. Clearly, this implies that various aspects of education for sustainable development are likely to be interconnected.

2.3. Course Organization and the Nudge Experiments

The two faculty members on this course had followed a broader agenda for working with local stakeholders on projects relevant both to students and to the organizations for several years [13]. The Qatar Behavioural Insights Unit was helping to set up nudge units in a few other MENA countries and had started working with the Kuwait Policy Appraisal Lab and the UNDP office in Kuwait. Part of the strategy for making the local nudge unit sustainable was to make sure that there would be a few nudge-capable graduates coming into the workforce. Hence, they had developed a nudge course and worked with a local university to deliver it as they went in to set up the nudge unit. We were approached to implement the course, with co-teaching by QBIU, KPAL, and the UNDP office every week during the first part of the course.

The intensive teaching segment of the course thoroughly grounded students in nudge theory. Students made a number of presentations about how the theory actually translated into real-world behavior for a variety of sustainability issues, particularly focusing on how a specific subconscious bias could vary depending on the nature of the action we wanted to nudge people toward. This was essentially the “head” part the schema in Sipos et al. [51], learning the theoretical concepts and what they mean beyond just the pages of the book. During this phase, students started brainstorming on what kind of small project they could do that would have practical value to a client but would also be small enough that they could implement it with very limited time and money budgets. The presentations gradually shifted to examples related specifically to their short-list of projects.

By late in the first half of the course, consultations with the nudge experts and initial contact with management at a few short-listed projects had identified which project they would do. Attention shifted to choosing and negotiating implementation of the actual real-world experiments (“hand” in the Sipos et al. [51] model). The trash behavior projects were considered the most feasible topics, which could give extensive data in a short time so that students and management could actually see the impact of nudges. One project worked with a local cinema, aiming to get patrons to increase the use of trash bins, to reduce clean-up costs between movies. The clean-up is labor intensive, because it requires a number of janitorial staff to be fast, so it can be done before the next screening starts. The other project aimed to get better sorting at recycling bins on campus. The decisions included the consideration of how committed the client organization would be in implementing the nudge.

Commitment did not constitute much money, but it did require a certain amount of time to work out feasible implementation details and authorization to perform the actions needed. The nudges were not very intrusive, so there was not usually much resistance, but occasionally, some specific element would not work within the organization’s system and constraints. For example, an unobtrusive reminder message immediately before a behavior takes place is often an effective nudge. However, contracts with movie suppliers do not allow the cinema to insert a message at the end of the movie before the credits, because the credits must appear with the movie. By the end of the credits, the message would be legal, but most people would have gotten up and left by then, so it would not be effective. A different plan was needed. Students worked closely with management in each location to plan every detail of the interventions and to make sure that collected data could help management judge cost efficiency. In both cases, management funded project material costs, but of course, nudges aim to be cheap, so these costs were not very great. If the nudge worked, they could use it regularly, with the aim being to reduce costs more than those that were originally invested in the nudge.

2.3.1. Experiment 1 Cinema

Students spent considerable time at the cinema observing traffic flow and talking to patrons to work out an effective intervention. They identified several key issues: 1, patrons might intend to put trash in the lobby bins, but simply forgot; 2, perceptions that it is the cleaners’ job; 3, lobby bins are not visible from inside when people get ready to leave; and 4, bandwagon effect because so many other people did not do it. Nudging experts are well aware of such subconscious System 1 forms of bias, and the student teams worked out several reminders to nudge people toward more bin use. Various forms of reminder nudges can be effective for many of the reasons.

The students created short audio videos shown right before the film started, with a logo backdrop. Figure 2 shows the still logo, while the short repeating loop with these graphics was animated. The audio stated (in Arabic): “Dear guests we’re counting on you to keep the theater clean so please remember to throw your trash in the bins at the end of the movie”. This primed people, but by the end of the movie, they might have forgotten, so the still logo was also attached on the back of seats. To avoid distracting viewers, the color scheme was difficult to see during the film, but easy to see on the seat in front of viewers when the lights came on. To illustrate the amount of detail that went into planning, it is instructive to mention that management was concerned that the label adhesives might leave residue and damage the seat material, so various adhesives were tested first to avoid this, while also making sure that the label would not fall off.

The cinema had six theaters and every day normally showed six films in each, taking up the whole day. There were three ratings, roughly general audience, adult supervision, and adults. The experiment lasted nine days, and a total of 73 shows were analyzed, 49 with the treatment and 24 without, randomized by film, its rating, and time of day. Practitioner-oriented researchers often recommend that it is more important to understand the treatment response than where nothing is done in the control. This must be balanced to avoid sacrificing much statistical power, which is usually maximized with equal proportions of control vs. treatment. White [62] demonstrates that power declines only slightly with the control proportion, down to about 30 percent, but then declines rapidly for lower control proportions.

The main data were from students working with the janitorial staff to count and weigh all the trash in the bins and on the floor after each show. The weighing was an innovation that had to be decided ad hoc, because students learned quickly that there was not always enough time to count during short between-film times. Quick adaptation in the field is an important lesson for students. Malodia et al. [63], for example, note that marketing research field experiments must remain flexible to deal with situational changes. The student team also had to coordinate tasks, not only for data collection (counting/weighting trash from the bins and from the floor), but also to check the back-of-seat logos before each screening, and to oversee janitorial staff to make sure bags of floor trash did not get mixed with bags from the bins.

Overall, the treatment was effective and increased trash in the bins from 30.84% in the control shows to 49.45% in the treatment shows (

Table 1. Percent trash in bins, basic ANOVA statistics for treatment and rating.

	Overall Mean	std dev	n
treatment factor
control	30.85	17.993	20
treatment	49.45	16.019	49
ANOVA sig. treatment	0.000
rating factor
E	41.10	16.548	26
PG	36.80	23.351	15
T	50.70	15.695	28
ANOVA sig. rating	0.001
ANOVA sig. interaction (treatment × rating)	0.009
overall ANOVA model sig.	0.000
ANOVA R-sq	0.386

Note: four control cases were slightly contaminated and are not used here. This was only a few trash bags out of a large total, and the cases fit the same pattern as other control cases. However, it was judged best not to introduce potential slight bias.

). This differed quite substantially across ratings levels (which attract different audiences), but for all levels, even where the bin usage was higher than for other ratings, the treatment increased bin usage (Figure 3). These results are for the overall percentage of trash in the bins after the show. As a check, the same analysis was performed for the actual weight of trash in the bin on an individual level (total weight/number of viewers). The pattern was very similar.

Such results suggest the issue discussed above—sometimes attitudes are quite weak and do not translate into action well [45]. The context (here, nature of the film and audience composition) can itself have an impact on raising subconscious System 1 thinking to more conscious System 2, which can shift behavior toward the better usage of trash bins even without nudging. But even the rating that gained the best bin usage in the control group showed a small benefit from nudging. In the other contexts, which had much less bin usage, the nudge impact was much stronger, so that the end result was bin usage at fairly similar higher levels across the three ratings/audiences.

2.3.2. Experiment 2

This experiment aimed at improving trash sorting by the university population, into bins for plastic, paper, aluminum, and other. Each of the two main bin locations has four bins, with originally small, fairly ineffective labeling as to which sort of trash should be put in each bin. The A bins where the intervention was implemented are between several classroom buildings, which have coffee shops and canteens, and B bins are near the admin building, library, and bookstore, where there is little food service or food retail availability. There is substantial traffic past B, but not as much as at A. The intervention was carried out only on the A bins. But the campus is relatively small, and many people frequently pass both locations, so B bins could not be used as a control.

Two different treatments were used. The first was a combination of more prominent and visually appealing stickers on the bins, to call attention to the materials that should be sorted into them and removal of the lids so that people did not have to open the bins to place things in them. The first element in this combination is basically a reminder nudge, as in the cinema case. The second element is a simple infrastructure modification to remove a minor barrier, although not as big a barrier as the absence of infrastructure noted above in [5]. The main color on the stickers represented each material, according to a week of prior observation to determine what color much of each material typically was (orange for plastic, green for paper, blue for aluminum, and black for other general waste—this is related to colors of popular brands in the stores on campus and cannot be generalized). The second intervention added a hoop over the bins to turn it into a simple basketball shot. The stickers remained, and the lids remained off.

Most people are on campus several days a week, so randomization by day would not work because once exposed to the treatment, thinking would already be affected on returning to a no-intervention day. The experiment lasted for 24 days, and the first eight days served as the control, then eight days for treatment 1, and finally eight days for treatment 2, which was the hoop added on to treatment 1.

Table 2. Percent correct material in bins, basic ANOVA statistics for treatment and location.

	Overall Mean	std dev	n
treatment factor
control	26.07	0.303	54
treatment1	49.06	0.331	56
treatment2	50.67	0.276	43
ANOVA sig. treatment	0.000
location factor
A	46.99	0.290	93
B	32.74	0.359	60
ANOVA sig. rating	0.010
ANOVA sig. interaction   (treatment × location)	0.294
overall ANOVA model sig.	0.000
Anova R-sq	0.171

Note: n is not the number of days, because of multiple collection times each day.

summarizes the ANOVA results. Figure 4 indicates that there was a very large increase in correctly sorted trash at location A. During the control days, 26.6 percent of trash was correctly sorted overall, nearly equal in the two locations (A 27.2%, B 24.7%). At location A, where the treatment was carried out, this rose to 56.3 percent during treatment 1 days. Adding the hoop did not change this, and correct sorting remained about the same, 56.6 percent during treatment 2 days. In other words, the nudge doubled correct trash sorting, but two nudges at once did not perform any better than just the first one. It was quite interesting, however, to observe location B, where there was no nudge. Much of the traffic on campus would pass both locations in a given day, and correct sorting also rose, to 39.5 percent correct sorting during treatment 1 and 35.4 percent during treatment 2. Overall, correct sorting for the combined two locations rose to 49.1 percent on treatment 1 days and 50.7 percent during treatment 2 days. These overall figures reflect the heavier weight on data at location A data, since it was in a location that generated more trash.

Results at location B, where there was no treatment, are particularly relevant to the discussion above about shifting low-involvement habits [39]. The nudges clearly work at getting more sorting compliance in location A. Whether they actually bring either knowledge of recycling issues or attention to recycling actions in Figure 1 above to the conscious System 2 level is open to debate. Location B results, however, suggest that habits, even if subconscious System 1, can be influenced even in locations where there is no nudge if people have experienced the nudge elsewhere. Location B also clearly showed an increase in correctly sorted trash. It was not as large as at A, but people carried their improved sorting behavior over to the other location to some extent. This may still be at the subconscious System 1 level and may fade if nudges about sorting are discontinued so that people do not experience them anywhere. But the point in [39] is that subconscious System 1 habits can be changed into beneficial habits. Location B here suggests that long-term visible nudges in one location would be able to establish the level 4 (in Figure 1) beneficial habits in other locations where many of the same people go, even if there are no nudges in those other locations.

3. Discussion

A decade and a half after Sipos et al. [51] criticized traditional educational methods and proposed their head, hands, heart schema, Claro and Esteves ([9], p. 1249) still feel the need to argue that “embedding sustainability issues into education goes beyond merely presenting definitions and issues of relevance to business. It is also about developing several key skills that managers and managers-to-be can put into practice”. This suggests that despite the substantial increase in academic publications about service learning [49], academia has not fully embraced it, at least as applied to sustainability. Thus, there is substantial value in empirically demonstrating, once again, that approaches, such as service learning, work well.

The first lesson from this course is simply that nudging works—it improved consumer public trash behavior. Cinema patrons in experiment 1 increased the amount of trash they put in the bins from just over 30 percent to nearly 50 percent. Campus traffic increased the amount of properly sorted trash in public bins from 25 percent to 50 percent, and this changed behavior carried over even to other locations on campus where there was no nudge intervention. Nudge researchers frequently point out that nudge effects may be quite small and sometimes not statistically significant (quite true, as briefly discussed, e.g., in [41]). Here, the effects are relatively large, and they showed strong statistical significance. Simply in terms of student learning, it is simple for students to see the impact using standard basic statistical techniques that they have learned in earlier courses. There is no need to explain the nuances of looking more carefully, such as using non-parametric statistics to examine patterns of individual small, non-significant effects (as in [41]).

Of course, even relatively large nudges do not entirely eliminate problems and because “nudges do not always create large absolute shifts in behavior, scholars and policymakers may underappreciate their value” ([64], p. 1042). Here, however, in both of these experiments, improvement was substantial in getting more use of bins for trash and in correct sorting. Further, experiment 2 indicated that a nudge can change behavior beyond the specific location of the nudge (as at the B bins). Operationally, organizations can have an impact on managing consumer public trash behavior. This is especially important for small organizations. Nudges may not cost much, but small organizations operate at small volumes, and very small impacts may not be worth even a relatively small investment. The results here indicate that such nudges can be large enough to reduce consumer waste management costs even for small organizations.

Nudge proponents know this basic finding, of course; there has been quite a large amount of nudge work over the past decade, and the concepts and methods are well established (e.g., [65,66]). The more interesting aspect is that students designed and implemented these nudge projects. Nudge experts and faculty presented the basic theory and many examples and critiqued student thinking as they developed their understanding of nudge theory, the “head” knowledge base in building sustainability competencies, in the widely cited schema of Sipos et al. [51]. The experts remained available for consulting, and many classes in the mid part of the term were a discussion of implementation details that students were working out in coordination with clients. But initial qualitative research on consumer thinking was by students, project decisions were by students, and consultation/negotiation with clients was by students. The experts did not intervene in the fieldwork. Students gained extensive practical implementation experience, the “hands” part of the competency-building schema.

This provided substantial experience in managing an actual real-world project, but the benefits go well beyond this. Sustainability competence includes attention to how average consumers think and to coordinating with client stakeholders. The student groups consulted extensively with management to develop feasible interventions. They learned the importance of making nudges fit consumer thinking, as well as how critical it is to make everything in the nudge compatible with client conditions. (Recall the tiny details of matching colors on the sorting bins with the most common brand colors of trash at the location, a little consumer detail, or of testing out adhesive backing on the seat stickers, a little client detail.)

The authors have long used such action/experiential learning methods to give students practice with real-world projects (e.g., [13,67]). Regarding teaching practical marketing research, it was already clear two decades ago that “this hands-on teaching approach can offer an effective model for further development of the MBA curriculum at the universities in Vietnam” ([67], p. 53). Of course, there have been many such calls over the years, and in sustainability specifically, there is growing consensus that education is most effective when it incorporates action learning and practical experience (e.g., [8]). But the recommendation in Claro and Esteves [9] that this sort of teaching/learning approach is needed in sustainability curriculums suggests that implementation is still scattered, not widespread.

Even beyond competency, this sort of service learning can foster a sustainability mindset, which Cripps and Smith [17] and Hermes and Rimanoczy [18] call the emotional and being aspect of sustainability, essentially “heart” in Sipos et al. [51]. The process of building this sustainability mindset includes the work with the various stakeholders [25,26]. Participation in teaching the course and guiding the projects by personnel from the Kuwait Policy Appraisal Lab, the Qatar Behavioural Insights Unit, and the UNDP office in Kuwait demonstrated to students that this was not just a little academic coursework requirement (or a big one—the projects were a large amount of work). The benefits were two-way, students learned, and project results were incorporated into policy development by these organizations, contributing to work in Kuwait and more widely in the Gulf. Cooperation with the local clients where the experiments were carried out demonstrated that little nudges can work and will be implemented by such clients when proven to work.

This all results in student enthusiasm, as they discover that they can do good work that is used to make some small impact. Knowledge that classes have relevance in real-world environmental policy and that they can have some role in shaping that policy is a strong motivator toward keeping sustainability issues in mind. This helps strengthen weak attitudes [45] toward more conscious-level action, as in Figure 1 [39]. Experiment 2 here also suggests that nudges can help establish new, more sustainability-oriented subconscious habits, level 4 in [39].

Finally, at the university level, Beynaghi et al. [68] say that experiential, collaborative, and community-based learning will be an important part of the potential evolution of universities into prominent drivers of sustainability. Cuesta-Claros et al. ([69], p. 532) identify a category of transformative universities in fostering sustainable development, “systematically combining research and societal engagement paradigms such as transdisciplinary research and service-learning”. This is part of growing recognition that sustainability is not simply a technical problem (e.g., [70]). We certainly need good technical work, but it is not sufficient alone. Nudging can help in matching up technical solutions with human behavior, and the research behind good nudge projects makes sure that the technical details are feasible in terms of the behavioral response.

We do not imply that the university where our experimental courses were implemented is among these transformative universities. Like a great many higher education institutions in much of the world, it is somewhat more traditional in its approach to education. But individual faculty often are able to set up more innovative classes. With enough such faculty and graduates who make some real local impact, many universities will recognize the benefits of a more transformative role. In Kuwait, for example, the fact that a Kuwait government unit and regional and international organizations wanted to introduce the class [11] give the service learning approach some credibility in work to shift higher education toward more of a transformative approach.

4. Conclusions

This discussion has indicated how a service learning course was integrated into the economics program at a university in Kuwait. The key objectives noted in the introduction go well beyond simply learning the course content. Service learning and most varieties of experiential learning give students real-world experience to see how the academic concepts actually work in practice. External clients certainly appreciate exploring ideas on how to reduce costs somewhat, in this case, for handling trash. The sustainability literature argues that such courses help build a sustainability mindset and competence. Organizations working on sustainability issues, such as the Kuwaiti nudge unit and the UNDP Kuwait office, appreciate such training—and this goes beyond simply approving that students learn sustainability issues well. They want to hire a few graduates who have gained practical experience at these specific skills.

Recommendations based on this discussion should be fairly apparent. For sustainability educators, a small set of courses constructed this way can contribute to tackling challenges facing the operationalization of sustainability practices at varying local levels. Cooperation with local policy-making organizations and local businesses and institutions turns such courses into a forum for discussing important sustainability problems. At the same time, the courses train future practitioners in how to at least partially address various sustainability problems and provide a skilled workforce for organizations that need such skills to address sustainability issues. For students, this provides an education that is integrative and distinctively multidisciplinary as it integrates real-world application into education. The students see that even small projects can help at the local level, and this can feed back into classroom discussion about the “wicked” nature of sustainability problems [34,35]. There is no grand solution that will fix everything, so small projects are an important part of improving things.

Action learning should be a component of every program, including aspects of sustainability. There is a range of effective experiential/action learning methods, but at least a few applied service learning courses should be among them. Outside involvement by government agencies, international organizations and NGOs are an essential element, to give students guidance on what policies are possible and simply to show them that their work will be used in the organizations’ policy development. Consumer research is an important part of sustainability training—interventions and policies must fit local conditions to have much likelihood of success. Extensive interaction with client organizations is a must.

Through all of this, students learn more than just theory and gain experience demonstrating how even small policy initiatives can contribute to sustainability. The organizations that they work with gain practical tools for improving sustainability while reducing operating costs somewhat. Policy-making organizations gain information on what works well or not so well. Basically, such courses are win–win for a wide range of stakeholders who must be concerned about sustainability.

Limitations should be mentioned, starting with the two field experiments themselves (a college campus and a local cinema theatre). These represent real-world data, but like any controlled experiment, they come with some limitations. One of those is the context-specific experimental design, which can represent a challenge for generalization of findings. The cinema, for example, showed differences in trash behavior depending on the audience composition, although the end result after the nudge brought behavior much more similar in using the bins more. Audience composition will differ across cinemas and across countries and cultures. Here, the nudge impact was relatively large, but Figure 1 indicates that it was much smaller for one audience type (although there was still a small impact). Similarly, the campus for the sorting intervention is relatively small. On a bigger campus, there might not be much transference of nudged behavior to other non-nudged locations. Further, Kuwait is mostly hot and dry, and there might be different behavior in a cooler, wetter location.

Such factors can make it more difficult to control some of the variables in the experiment and increase variability, which could pose a challenge to separating nudge effects of different interventions. Also, the quantity, quality and reliability of the data can be a concern due to the complexity and logistics of field experiments as they are not conducted under controlled environments. Finally, all of these can collectively hinder experiment replication and slow efforts to build and validate a body of evidence. The upside, however, is a growing mass of data that allows for more careful attention to how context can play a role in how well nudge interventions can work.

Future research should examine a few of these issues. For example, in line with the applied nature of this course, the different trash behavior of different audiences should be explored. In a broader sense, this is about differences in any small subconscious, habitual behavior, which has a sustainability impact. Here, the several audiences were all in the same cinema, so a single intervention targeted the range of trash behaviors and improved everyone’s behavior but with bigger impact on the worst pre-intervention behavior. In some situations, the several behaviors may be in separate locations, and specifically targeting an intervention on those worst behaviors first can result in bigger improvements. Nudges do not cost very much, but in cases of very limited resources, organizations can make a bigger impact by targeting one intervention carefully rather than two nudges for worst vs. simply “could-be-better” behaviors.

Similarly, an examination of context issues is needed, such as the specific example here of hot, dry Kuwait vs. a cooler, wetter location. This can even be structured to help internationalize the educational experience. De Wit and Altbach [71] discuss the prominent trend of “internationalization” nowadays as often being oriented toward competition to attract international students for revenue generation. This does not do much to help foster sustainability. They note, however, that “collaborative online international learning” is a minor trend, which allows jointly taught virtual courses and active student involvement from both countries. A comparison of country context in such a virtual service learning course would be relatively easy to build into the discussion, as well as such issues as somewhat different key sustainability issues across countries or differences in dealing with stakeholders. An added advantage of such a course would be bringing out the point that sustainability is an international problem, not likely to be solved purely on the national level.

Overall, then, such service learning courses have an important role to play in teaching sustainability. For students, they help develop sustainability knowledge (head), instill implementation competencies (hand), and foster an appreciation and pride that even their small student project made some contribution to sustainability (heart). For the university, a few service learning courses help forge ties with local businesses and other organizations and with governmental and NGO stakeholders working on sustainability issues. As noted above, this seems like win–win all around.