Introduction

The current level of resource extraction, consumption and throughput to waste generation is a serious and global environmental issue [1, 2]. More sustainable approaches to using and managing materials and resources, including reducing use, are becoming increasingly important [3, 4]. The concept of a circular economy is one solution that can reduce negative environmental impacts while benefiting economy and society [5,6,7]. In essence, a circular economy involves reframing the way resources are used and managed [8]. The transition to a circular economy will require significant change at all stages of the economy [9]. This is clearly illustrated by Selverfors et al. [4], who illustrate the new flows and activity required at each stage of a circular economy (shown in green in Fig. 1).

Fig. 1
figure 1

Traditional representation of strategies in a circular economy [4]

Full size image

As Selverfors et al. illustrate, the majority of these new circular flows are initiated at the downstream ‘use/consumption’ and ‘end of use/life’ stages. In many ways, circular flows are therefore dependent on users [4]. Facilitating such ‘demand side’ changes is an important pathway for a wide range of sustainability transitions, including mobility, energy, waste, food and the built environment [10,11,12]. In fact, many have gone so far as to say that the goals of a circular economy cannot be achieved without appropriate user involvement [see e.g. 9, 13, 14].

Users are important because demand side changes have interdependent feedbacks with the supply side. Improvements at the use stage of the value chain can be achieved by upstream changes, as the material intensity of consumed products is driven by global networks of production [14, 15]. Positively, user behaviour can also enable the adoption of transformational circular and sustainable business models by businesses and civil society organisations [16,17,18]. For instance, users can influence both the volume and type of products placed on the market through purchase decisions, and the flow of products downstream or back upstream through divestment and disposal decisions [3, 19, 20].

More problematically, issues at the use and end-of-use stages have already been identified as critically hindering the transition to more circular business models [21, 22]. Particularly, the demand side can be a critical source of inertia, perpetuating socio-technical systems that make unsustainable behaviours the ‘path of least resistance’ [23, 24]. In addition, consumer behaviour can undermine upstream achievements through moral budgeting, negative spillover and rebound effectsFootnote 1 [25,26,27]. For example, experimental research has found that the availability of recycling increases resource consumption [28, 29], while the ability to donate unwanted clothes can lead to rationalising of increased new purchases [30]. Similarly, the ability to purchase cheap, second-hand clothes through thrift stores can make more income available for additional second-hand or new purchases [31].

Despite the importance of the users’ actions at the use stage, most circular economy literature to date has been heavily focused on the upstream production/supply side of the economy, with attention directed at the role of business/industry and policy to support them [32,33,34]. In fact, a recent review of circular economy definitions in the literature found that only one-third included aspects of consumption [35].The importance, function and enablement of users has only recently begun to receive attention [see e.g., 9, 13, 19].

Existing research on users and consumption in a circular economy has primarily focused on identifying factors that enable or hinder circular consumption behaviour, and the nature and dynamics of these practices, as well as the user position from a design perspective [see 21 for a review]. Much of this research has been conducted without a clear and comprehensive understanding of what types of new or altered behaviours will be required by users. For example, Corsini et al. [3] report that typologies of behaviour are rare in analysis of electrical and electronic equipment consumption. This is despite a number of authors stating the field needs to better identify user behaviours in order to understand the key barriers and enablers and attempt to shift them [8, 19, 21].

The aim of this work was, therefore, to produce a comprehensive framework clearly presenting the range and variation in behaviours required of users in a circular economy. We begin by considering the benefits of a behavioural approach, and current attempts to define circular consumption behaviours. We then outline our iterative and collaborative approach to developing such a framework, and present the results of this process, including the final Framework. This is followed by a discussion of the insights that arose from the process, along with implications for research and practice, including potential directions for future iterations of the Framework.

The Benefits of a Behavioural Approach

The behavioural science field generally posits that the first step in understanding and changing behaviour is to clearly identify, define and prioritise behaviours of interest [36,37,38]. This step is critical because even slight variations in behaviour, such as different target products [3] or contexts [19], can have different drivers and barriers, and require different interventions. The step of clearly defining behaviour is often overlooked in social and behaviour change research [39], including, as noted above, in the circular consumption literature. A ‘behaviour’ can be clearly defined by specifying the physical actions (action), performed by certain individuals, or groups of individuals (actor), often towards something or someone (target), in a particular time (time) and place (context) [39]. A clearly defined behaviour usually involves actions that are observable, and specified at a level where they are indivisible from other actions [36]. Furthermore, ‘end-state’ behaviours are preferred, rather than preliminary or precursory behaviours that simply enable a behaviour of interest [40].

Seven existing papers were identified by the authors that do attempt to summarise types of user/consumer behaviours. These papers discuss ‘behaviours’ in terms of broad categories [4, 8, 19], the different ‘roles’ that consumers/user may adopt [9, 34], or types of circular consumption ‘strategies’ [32, 41]. These papers do not, however, adopt a specific behavioural lens. For instance, they do not always focus on observable actions. Non-behavioural elements exist, including ‘product attachment’ [8], ‘reducing operating costs’ [8] and ‘rewarding circular business models’ [42], as well as enabling behaviours, such as ‘registering device’ [8], ‘separating waste’ [19] and ‘making use of product labels’ [42]. The papers also rarely specify behaviours at a level indivisible from other behaviours, instead utilising broader categories, including ‘optimising product choice’ [42], ‘product care’ [8], ‘share products and services’ [19] and ‘re-buy used products’ [41]. And where the action is clearly defined, only limited sets of behaviours are included [4, 32].

A specifically behavioural approach to developing a comprehensive typology of user circular behaviours could add much value to the literature. Firstly, it could provide both broader and deeper coverage of the full spectrum of actions that are desired or required to enable a circular transition. This in turn would allow the field to more specifically investigate and distinguish the drivers and barriers required for different actions, allowing more nuanced and targeted development of necessary enablers and interventions aimed at users [36, 38, 43]. As such, it would also have immediate application to policy and practice, which nearly always ultimately aim to change behaviour. Secondly, a comprehensive typology could enable more precise understandings of the interrelationships between user circular behaviours, and up- and downstream activities, including production and collection. This could spark innovation in both research and practice by highlighting new/previously unexplored opportunities for circular business models and offerings [33]. Thirdly, clearly and precisely defining behaviours without losing connection to context (the material life cycle in this case) has several advantages in terms of facilitating inter- and transdisciplinary research. Kaufman et al. [44] have suggested that focusing on actions/behaviours in context can be ‘a-theoretical’ and ‘a-disciplinary’ to the extent that a wide range of perspectives and disciplines touch on if not focus on behaviour. In doing so, behaviour can be understood as an empirical ‘boundary’ object over which different disciplines, sectors and jurisdictions can intersect and converse—even if they do not agree (deliberative dialogue). In doing so, it also provides diverse perspectives and approaches a consistent ‘point of entry’ into the multiple scales within the socio-technical systems shaping circular economy initiatives.

These benefits of a behavioural lens are particularly important given the problematic characteristics of circular economy research and scholarship outlined above and those we discuss in the next section. We believe circular economy needs to be addressed as a transdisciplinary topic, transcending boundaries both within academia (i.e. between disciplines) and beyond, with policymakers, industry and civil society. Some work is already collaborative and practice-focused, notably Selvefors et al. [4], but further, similar efforts are needed.

Current Understandings of User Circular Behaviours

As Milios [34] observes, the terms we use affect our thinking, and so we begin this section with some notes on language, before exploring existing attempts to define user circular behaviours. Firstly, we note that ‘circular economy’ is a heterogeneous, and umbrella term [13] bringing together a number of different ideas that centre around reframing the way resources are used and managed [8]. It has been defined in academia in many ways [35, 45] as well as by different societal actors such as government, business and civil society [13]. For the purposes of this study, we consider circular economy from a ‘technical materials’ perspective [46, 47], and see the key aspects of the concept to include: transforming systems of production and consumption [21, 48]; minimising resource inputs and system outputs/leakage [49]; and keeping products, components and materials circulating at their highest value for as long as possible [50, 51]. Key principles include narrowing, slowing, intensifying and closing [52, 53]. To further clarify the scope of the concept, we note that single-use and disposable items are problematic in the circular economy, and should be ‘designed out’ wherever this is possible. This means they are not centrallyFootnote 2 part of the conception of circular ‘products’ [32]. Thus generally, the ‘target’ items of circular behaviours will be durable products that retain their function and form during ongoing use, and which are never completely ‘used up’ though they may become ‘worn out’, broken or obsolete [54].

Secondly, in specifying the general ‘actor’, we note some diversity in terms. A very common term in existing studies is ‘consumer’ [13, 41, 42] linked to the ‘consumption’ phase of a traditional product lifecycle. However, ‘consumer’ is problematic for three reasons. Firstly, as noted by many before, the term implies a passive role for this group of actors [34, 55]. Secondly, as Lindahl and Dalhammar [56] highlight, we should not continue to use terms that represent old ways of thinking. In this regard, ‘consumer’ has a strong alignment to existing patterns of linear consumption of resources, from which a circular economy is transitioning away from. Finally, in policy/practice, the term ‘consumer’ often implies individuals operating in/for their personal or ‘private sphere’ [57], in contrast to operating as or on behalf of business/industry, civil society or government entities [58]. However, products are utilised in all of these spheres, so circular behaviours at the use stage are pertinent to actors across these public, commercial and civil spheres as well.

One alternative term that rejects passivity is the (circular) ‘citizen’ [34, 55]. We acknowledge the value of this term, particularly the sense of responsibility and agency implied. However, we note that the term citizen also denotes an ‘individual’ sphere that excludes some types of roles that individuals who utilise products adopt, such as bureaucrat, employee and business/industry leader, and does not, therefore, encompass all of the ‘use’ stage. Another term frequently adopted is ‘User’ [4, 8, 33], which is flexible across private, public, commercial and civil spheres. It aligns well to a technical materials/products scope, though we acknowledge it could exclude consideration of important broader, enabling behaviours around advocacy and system change.

Regarding specific user circular behaviours, as noted above, a number of previous papers have attempted to summarise the various behaviours or roles of users/consumers. Five of the seven studies we identified distinguished different phases of consumption or user interaction with products, and used these to organise included behaviours. As shown in Table 1, all utilised three phases, which, despite differing terminology, can be considered mostly analogous across all five studies. Such a tri-part distinction is also found in other circular consumption research [20].

Table 1 The three phases of User interaction with products
Full size table

All seven focused on the ‘action’ aspect of behaviour, with occasional references to targets, contexts and timing. A visual comparison of the seven efforts is shown in Fig. 2. It reveals substantial divergence in the phase of product lifecycle covered, the types of behaviours included, the specificity at which behaviours are defined and the language adopted to label these behaviours.

Fig. 2
figure 2

Current types of user actions identified in circular economy literature. Note: Horizontal alignment represents alignment between definitions of behaviours. Different areas of overlap indicate subtle differences between definitions of similar behaviours. The number of light grey squares in each column represents the number of missing behaviours in each paper, compared to the full set of behaviours covered collectively

Full size image

Maitre-Ekern and Dalhammar [42] had the greatest coverage range and fewest gaps across the full set of behaviours from the seven studies. Despite this greater coverage, they noted that their specified roles and activities were likely to continue to evolve and require further mapping in the future, suggesting the need to continue to iterate frameworks of consumer behaviour. Considering specificity, some studies used general categories like ‘repair’ behaviour [8, 41] while others distinguished the specific action such as user ‘DIY’ repair or community repair as opposed to commercial repair [9, 32]. Both Wastling et al. [8] and Selvefors et al. [4] also distinguished between behaviours involving user ownership and those based around access, while Selvefors et al. differentiated between behaviours involving users and businesses, and those involving users and other users.

In terms of language, in a number of cases, the same or very similar behaviours were differently labelled. To illustrate, at the acquisition stage under avoiding purchasing unnecessary items, Maitre-Ekern and Dalhammar [42] list ‘Avoid replacing products that work’, while Camacho-Otero et al. [41] and Chamberlin [32] list ‘Retain items’, clarified by Camacho-Otero et al. as ‘not disposing prematurely’, while Wastling et al. [8] list ‘Postpone / prolong replacement’. In other cases, the same labels were defined, used or classified somewhat differently. For example, Chamberlin [32] lists ‘Renumerate’ defined as paying more at the acquisition stage for an item or service with greater longevity, while Camacho-Otero et al. [41] define renumerate as getting additional monetary value from the product during the use stage. While not further clarified, we believe this implies behaviours such as lending through paid peer-to-peer sharing schemes. This lack of clarity in definitions of terms was apparent in a number of the studies, making it a laborious and likely imperfect endeavour to match different terminology across the studies to produce Fig. 2.

Finally, an important point made by Maitre-Ekern and Dalhammar [42] is that some behaviours can have negative environmental impacts in some circumstances. Circular benefits only arise when these behaviours are chosen instead of ‘linear’ behaviours, a concept known as replacement or displacement [see, e.g. 60, 61]. This point is central to the European Environment Agency (EEA) typology [59], which pairs circular behaviours with opposing linear behaviour based on the decision(s) that prompt the behaviours.

Method

The Framework presented and discussed in this paper is the outcome of an iterative and collaborative process inspired by Selvefors et al. [4]. Development of the current typology involved input from a wide range of academics, policy-makers and practitioners over three years through a series of workshops and asynchronous activities, supported by a number of Australian federal and state governments. The initial two phases were governed by the needs of the different sponsoring government stakeholders via an existing cross jurisdictional, academic-policy research collaboration involving both authors [60, 61]. The final phase was devised by the lead author as part of a larger project funded by the Australian federal government [62]. This phase adopted a much broader engagement approach as the backbone of an explicitly transdisciplinary and interdisciplinary research integration effort. Research integration aims to improve understandings of real-world problems by synthesising relevant knowledge from a diverse range of stakeholders including academics and non-academics, particularly those in a position to make decisions about the problem [63,64,65]. The full process is summarised in Fig. 3 and described below.

Fig. 3
figure 3

Process for developing Framework of User Circular Behaviours

Full size image

The process was initiated in 2019 by the Australian federal government, to develop and prioritise an inventory of household waste prevention and management behaviours in support of the cross jurisdictional research collaboration. This first stage of the work began with the lead author conducting an exploratory review of academic literature seeking existing reviews that identified, defined and/or categorised relevant behaviours. Few were available at the time, and so the search was supplemented with a desktop review of grey literature, particularly government and environmental civil organisations’ websites that promoted circular economy, or waste prevention. These websites provided lists of target waste prevention/waste management behaviours, which were compiled into a ‘long list’ of potential behaviours, organised according to the three main stages of consumption identified above (acquisition, consumption/use, end-of-use/life) with sub-categories inductively developed by the lead author. The list and categories were refined in a workshop with the two authors, two additional behavioural science researchers and fourteen federal and state government policy makers, into an initial draft of behaviours and categories. The workshop also identified the need to expand the list to include innovative household circular behaviours not present in existing waste prevention frameworks, in order to align with emerging circular economy policy priorities [60].

The next stage therefore entailed a series of synchronous and asynchronous activities in 2020 amongst a subset of the previous government organisations, specifically the two authors and seven representatives of the Victorian State Government. The list of behaviours was interrogated and expanded to match the policy goals and activities outlined in Victoria’s new circular economy policy [66]. The categorisation of activities was also refined by the lead author through another exploratory literature review focused on identifying different types of activities and ways of classifying them, resulting in a Preliminary Framework [61]. This was then tested by exposing to the broader group of federal and state policy-makers. Feedback received centred around issues of clarity and the non-universality of language; e.g. the same terms meant different things to different stakeholders.

Consequently, the third and final stage took the Preliminary Framework into an extended consultation process throughout 2021–2022 with a range of Australian circular economy experts drawn from the public, private, research and civil society sectors. Key stakeholders were collated by the lead author through a combination of existing knowledge of the research team/project steering committee, desktop review for relevant organisations and authors, and snowballing via recommendations from existing invitees. Stakeholders were invited to participate in one of two workshops or to provide feedback offline on workshop inputs. All participants were then sent a preliminary draft for final comment, which was also released publicly for feedback from stakeholders not explicitly identified/invited. In total, 94 stakeholders engaged over a 6-month process from 38 organisations, as outlined in Table 2 [67].

Table 2 Number and type of stakeholders involved in final phase of collaboration/research
Full size table

The workshop process focused, firstly, on clearly distinguishing behaviours through precise definitions and descriptive labels, which involved dialogue, negotiation and compromise amongst stakeholders to reach final agreement. Such conversation-based integration of expertise from the many perspectives represented is a key aspect of transdisciplinary research on real-world problems [63]. As part of this, some preliminary consideration was given for how the user behaviours interacted with up- and downstream actors, and the enabling behaviours that were therefore required both upstream and downstream. The second part of the process involved grouping behaviours based on their ‘function’ or purpose within a circular economy. The emerging set of hierarchical categories became the final organising structure for framework. In total, 112 Australian stakeholders collaborated with the authors across the three stages to co-produce the final Framework of User circular behaviours [62].

The final Framework was then utilised by the authors in two ways. Firstly, the set of user behaviours was compared to existing sets of behaviours in the international academic literature to understand similarities and differences. Secondly, the initial thinking on up- and downstream activities and their interactions with user behaviours was further refined into two outputs setting out the dependencies between user behaviour and circular activity up- and downstream.

Results

The following sections provide a summary of the dialogue and decisions made collaboratively through the three stages of the stakeholder collaboration process, along with the key outputs throughout the process. First presented is the organising structure adopted, followed by the full set of actions in the new Framework of User Circular Behaviours (‘the Framework’) and a comparison with existing work. The final aspect presented is the interrelationships between user behaviour and the behaviour of other actors working up- or downstream to implement the circular economy.

Organising User Circular Behaviours

In stage 1 of the process, the three phases of user interaction with products (acquisition, use and end-of-use/life [4, 32]) were used to organise the list of behaviours collated. In stage 2, these categories were identified as both too broad (not enough differentiation) and too vague (not descriptive enough) for the purpose of making sense of the many behaviours included. In addition, the ‘Acquisition’ category was not considered universally useful as the first stage, given that certain circular behaviours actually involved reducing consumption by avoiding acquisition, a category akin the ‘R0 Refuse’ strategy [68]. Further review of the literature also revealed some useful distinctions between different potential purposes of behaviours at the use and end-of-use stages. A broader set of categories was therefore developed, drawing particularly on the work of Glover [54] to distinguish retainment behaviours from consumption/use behaviours, and similarly distinguish divestment behaviours from discard behaviours, distinctions echoed by Reike et al. [68] and Camacho-Otero et al. [41], respectively. These preliminary categories are shown in Fig. 4.

Fig. 4
figure 4

Preliminary categories for organising user behaviours

Full size image

During the third stage, these preliminary categories were interrogated and considered generally useful for organising the behaviours. However, it was decided to relabel the six overarching categories based on their function in a circular economy, to better understand what makes each of the included behaviours ‘circular’. Inductive grouping exercises by stakeholders also produced a set of final sub-categories, or ‘types’ of behaviour within the higher-order functions, to further distinguish the behaviours.

The final organising structure for the Framework is shown in Fig. 5. At the top level, it adopts the three previously discussed phases of a user’s interaction with an item to organise the Framework, plainly labelled as ‘Get’, ‘Use’ and ‘Pass on’. The second layer of categories describes the circular function of included behaviours, while the final layer of categories distinguishes the different types of behaviours.

Fig. 5
figure 5

Structure of the Framework of User Circular Behaviours

Full size image

During final discussions of the organising framework, stakeholders noted that, while not deliberately intended, the various functions closely aligned to the four core principles of a circular economy (narrowing, intensifying, slowing and closing) noted earlier. This alignment with circular principles is also indicated in Fig. 5.

As part of consideration of the final organising structure, stakeholders also discussed whether the organising structure conveyed an explicit or implicit priority order amongst actions. After extensive dialogue, two conclusions were reached. Firstly, the order of functions was considered to represent an appropriate priority order, as per traditional ‘waste hierarchies’ where the upper (or in this case, earlier) actions provide greater environmental benefits than lower (or later) actions [42]. For example, it was agreed that, generally speaking, avoiding the need for an item would provide greater environmental benefit than accessing an item without taking ownership, which in turn has greater benefit than purchasing a second-hand item, or a new item made from circular materials. Similarly, avoiding an item entirely or continuing to use an item and investing to retain its value will generally provide greater environmental benefit than simply passing on acquired items whenever no longer wanted. Secondly, a concomitant conclusion was that any ‘hierarchy’ is a guiding principle only. Determining which behaviour is the most beneficial and appropriate in any circumstance will always depend on the individual product and context. An example given was older fridges, where replacing with newer fridges containing more efficient technology might provide greater environmental benefit than prolonging the working life as long as people. Furthermore, such determination would need to consider factors beyond material impact, such as energy use and associated emissions, and therefore lifecycle research is required to formally quantify and rank the benefits of behaviours under specific conditions.

Framework of User Circular Behaviours

The core product of this process was a comprehensive list of user circular behaviours. Given the level of detail involved, it was agreed early on to focus only on specifying the various actions involved. In some cases, the target and/or context are implied in the action, while in other cases, the action is flexible across various targets and contexts. The ‘actor’ was defined generally as any individual who ‘uses’ products or services. This was agreed to include individuals/households in their personal sphere, individuals working in/for businesses, including those businesses that also produce, supply or recirculate items and individuals working in/for other organisations that govern or otherwise support society.

The initial list of actions produced in the first stage of research from desktop review went through only small changes in overall composition in the subsequent two stages, mostly splitting ‘behaviours’ into more specific actions and, where necessary, specifying channels that clearly differentiated the action. The main transformations centred around the language used to label and clearly define the behaviours. This reflected a key finding throughout all stages: that different stakeholders often use different language to refer to the same concepts, or the same language to refer to differing concepts. An important principle agreed by all stakeholders including policy-makers, circular economy practitioners and industry representatives was the need for clear and precise language. This resulted in a deliberate turn away from the ‘Re-’ approach to labelling evidenced in 3-10R frameworks [68], as these were considered to result in vague labels open to various interpretations, as experienced by stakeholders in the workshops.

The final set of actions agreed is shown in Fig. 6, Fig. 7 and Fig. 8 broken down by the three main phases (Get, Use, Pass On). The complete Framework, incorporating precise definitions of each behaviour, is provided in Table S1 in Supplementary Information.

Fig. 6
figure 6

User circular behaviours within a material circular economy: GET phase

Full size image
Fig. 7
figure 7

User circular behaviours within a material circular economy: USE phase

Full size image
Fig. 8
figure 8

User circular behaviours within a material circular economy: PASS ON phase

Full size image

Comparison with Existing Frameworks

Our comparison of the set of actions in this Framework of User Circular Behaviours with existing sets of behaviours in the international academic literature is provided in Fig. 9. This includes the seven previously identified studies, as well as the 10Rs Framework [68], which is presented as a final, comprehensive set of circular ‘R’ strategies. The table illustrates that our Framework echoes elements of each of the existing works, but is both broader (contains greater array of behaviours) and deeper (contains more precise distinctions between behaviours) than the existing eight approaches identified.

Fig. 9
figure 9

Comparison of framework with existing attempts Bold indicates behaviour explicitly included in a 10Rs strategy. Asterisk (*) indicates behaviour implicitly included in a 10Rs strategy. Brackets ([]) indicate indirect involvement of user in a 10Rs strategy. Dash (-) indicates action is outside existing 10Rs strategies

Full size image

In terms of breadth, the current framework contains three specific actions that were not covered in any of the previous studies: make the item (DIY), repurpose the item and upcycle the item. These three behaviours are all ‘informal’ in the sense that they do not involve direct interaction with the formal economy, and this may potentially explain their absence from previous works. From the perspective of depth, there are a number of specific behaviours in the current Framework that are not distinguished explicitly in any of the previous studies. For example, the distinction in this Framework around borrowing, specifically between borrowing item from personal networks, borrowing item from community initiatives such as ‘libraries of things’ [69], borrowing item from commercial entities through traditional rental/hire arrangements and borrowing ‘access’ from commercial entities through product-as-a-service arrangements, is not reflected in previous works. Similarly, the various actions that enable sourcing an item second-hand, and ‘rehoming’ an item by divesting to another user are only partly distinguished by two, and three of the works respectively, and not fully in any of the seven.

The comparison also identified a number of inclusions in the previous seven works that were not included in this Framework, as shown in Table 3. A review of these missed ‘behaviours’ identifies that some are not observable actions but rather mindsets or outcomes of collective actions. Others are preliminary or enabling behaviours. Still others relate to broader circular aspects outside a technical materials perspective. Thus, we conclude that the collaborative stakeholder process resulted in a comprehensive set of user circular behaviours irrespective of these differences.

Table 3 Previously identified ‘circular user behaviours’ missing from the present Framework
Full size table

Inter-relationships with Up- and Downstream Activity

The final focus of the process was on how the specified user behaviours interact with activities upstream at the design, manufacture and distribution/retail stages and downstream at the collection and recirculation stages. This aspect was exclusive to the third stage of the process and was in response to participants’ assertion that user behaviours could not be clearly identified and defined in isolation from other parts of the value chain.

This process identified a complex web of interdependencies between users and other actors. Early discussions clearly recognised that some user activities are entirely contingent on choices made upstream at the design, production and/or retail stages (e.g. hiring items, buying refurbished items or buying items made from recycled materials), while others could be performed independently by users but could be made easier by upstream behaviours. For instance, repairing an item (self/third-party) can be done by ‘tinkering’, but is enabled by design for repair and provision of manuals, spare parts and tools. Similarly, onselling an item to another user can be done independently through traditional channels, but can be made easier by the adoption of digital product passports by producers/retailers. It can also be enabled by third-party activities, such as resale platforms that integrate digital passport information to store information on people’s collection of items and automatically populate ‘ads’ with pertinent information if they decide to onsell. Similarly, some user behaviours are contingent on downstream services like collection. A summary of these (in)dependencies of user behaviour on other actors’ behaviour is provided in Fig. 10. Specifically, it sets out dependencies with both up- and downstream actors, as well as other users, and third-parties such as service providers.

Fig. 10
figure 10

Relationship between user behaviours and up/downstream activities

Full size image

Another finding resulting from this discussion was that user behaviour could also enable or fulfil up- and downstream circular behaviours. For example, upstream behaviour like design for multiple use or provide end-of-use return can only achieve their inherent circular goals if the user ‘complies’ by reusing or returning the item; they are therefore contingent on the user. At the same time, some upstream behaviours create circular benefits without any direct action by the user, including designing efficient products or making with recycled materials. The list of up- and downstream behaviours and their dependency on user behaviour received only limited stakeholder discussion due to time constraints, and so was further developed by the authors.

The second output, in Fig. 11, is therefore a preliminary extension of the Framework to up- and downstream behaviours by actors involved in providing and/or recirculating products. The preliminary definitions of these behaviours are provided in Table S2 (Upstream) and Table S3 (Downstream) in Supplementary Information. In addition to showing dependency on user behaviour, Fig. 11, also sets out whether and how each of the up- and downstream behaviours are implicated in allowing/enabling user behaviour—the reverse of relationships shown in Fig. 10.

Fig. 11
figure 11

Relationship between up/downstream activities and user behaviours

Full size image

Discussion

The primary aim of this work was to develop a Framework presenting the full range and variation in behaviours required of users in a circular economy. Furthermore, we aimed to understand the relationships between user behaviours and other actors in a circular economy. Overall, our framework demonstrates that there are a large and diverse range of circular user behaviours, and these can be coherently and clearly described and grouped by their different functions and purposes. To our knowledge, this is the most comprehensive and detailed typology of user behaviours in a material circular economy produced so far. Furthermore, our analysis of the Framework within the broader circular economy system reveals strong interdependencies up- and downstream. We begin this discussion by further reflecting on these two findings before exploring some of the other insights that arose during development of the Framework. Implications for both future research and practice are considered throughout.

Greater Coverage and Specificity than Previous Attempts

The final Framework contains almost 60 distinct user circular behaviours. As noted earlier, a number of previous studies have explored the roles and behaviours of the user/consumer in a circular economy [4, 8, 19, 32, 41, 42, 59]. The comparison of this Framework against those previous works revealed much cross-over but also some gaps in the existing models. The various gaps in individual works likely reflect the specific aims and purposes of the previous studies, which generally included exploring particular aspects of user behaviour (such as those that support certain business models, e.g. Wastling et al. [8]; or those that occur between users, e.g. Selvefors et al. [4]) or summarising the ‘main’ types of user behaviour that are required for circular consumption (e.g. Camacho-Otero et al. [41]; Gomes et al. [19]). Compared to the existing body of works as a whole, the Framework provides only slightly greater coverage, indicating that most of the behaviours are being discussed to some degree in at least some of the relevant literature. It does, however, reveal some clear gaps in the consumer actions (both explicit and implied) in the ‘final’ 10Rs Framework proposed by Reike et al. [68], particularly around opting for access over ownership models, minimising purchase impacts by seeking items made from circular materials or with circular characteristics, and upgrading one’s own items to increase value/function to enable extension of useful working life. While the latter could easily be incorporated by extending the definition of ‘refurbish’ to include consumer action, the comparison suggests that further evolution of the 10Rs Framework is still needed to develop a ‘final’, comprehensive ‘R’ strategy framework.

Another clear difference between the current Framework and previous attempts is the level of specificity of behaviours and their labels. This is clear from both the comparison with the previous models, and with the 10Rs Framework. The specificity was an outcome of explicitly taking a behavioural lens, and seeking to define ‘observable’, ‘indivisible’ actions [36]. It resulted in an overall large number of behaviours, as well as many within individual sub-categories (such as ‘accessing item without ownership’, ‘sourcing items second-hand’, ‘passing on/divesting an item for use by another user’). The large number of behaviours highlights the ‘wickedness’ of the challenge of transitioning to a circular economy (see Fig. 12). At the same time, the Framework contributes to the solution by making it easier to break down the problem of user/consumer transition into specific behaviours, enabling these to be considered and targeted individually, simplifying change efforts [44]. Focusing on precise, ‘empirical’ behaviours (i.e. observable and measurable) reduces the need to adopt explicitly disciplinary conceptual ideas and theoretical constructs in primary research, while simultaneously providing boundary objects for such contributions to accrue and interact over. In addition, it avoids the confusion of behavioural antecedents or outcomes (for example, a sense of ownership) with the behaviours that they may relate to, i.e. maintaining or repairing a valued item. This facilitates both conceptual and practical clarity.

Fig. 12
figure 12

(Source: Smith & Kaufman, in Kaufman et al. [73])

Breaking complex problems down by identifying specific behaviours and audiences

Full size image

As noted earlier, the specificity of the Framework also facilitates behaviour prioritisation processes [36], and assists researchers, policy-makers and practitioners in correctly identifying the particular barriers and enablers of specific behaviours, rather than general barriers for vague ‘types’ of behaviours. The ability to accurate diagnose barriers and enablers is a necessary precondition for selecting or developing effective behaviour change interventions [74].

In terms of future research, this Framework provides two immediate opportunities to grow the field’s understanding of user behaviour in a circular economy. Firstly, the comprehensive set of behaviours contained in the Framework can be integrated into behavioural surveys and observation, and used as a starting point for measuring current adoption levels of user circular behaviours, in order to understand the extent of the transition still required. Secondly, the Framework can be used to accumulate and organise evidence of barriers and enablers of these behaviours, and the interventions that work to change them.

There are also opportunities to expand the coverage of the Framework. We note, for example, that our specification of the actor as a user of material products has constrained our focus to some extent. The focus on materials means this Framework currently excludes behaviours that seek to minimise use, and maximise value, of other resource inputs such as biological materials, water, energy and fuel. Our emphasis on ‘User’, even though extended to use of product and product services, obscures other types of behaviours unrelated to products, such as any ‘civic’ behaviours that may be required of ‘citizens’ in a circular economy. We note these additional types of behaviours as key avenues for future research and potential expansion/extrapolation of the current Framework.

Strong Interdependencies Up- and Downstream

Analysis of the inter-relationships between users and up- and downstream actors revealed strong interdependencies in achieving the circular benefits of particular behaviours. While literature has noted the importance of supply and demand side changes in transitioning to a circular economy (see introduction), discussion so far has remained relatively general. This new Framework provides a level of granularity and clarity to these relationships that was previous lacking. In addition to highlighting when user behaviour can be enabled by up- or downstream activities, it also emphasises that user behaviour is critical to the success of many up- and downstream circular endeavour.

The explicit mapping of interdependencies is a novel and valuable contribution to existing knowledge of the circular economy, and can inform both future research and practice. By applying the Framework to specific business models, products or material streams, actors in research, policy and practice can trace the thread of behaviours and their dependencies through the circular economy in a much more lucid, and comparable way than before. For instance, researchers and implementers of emerging circular business models could use the Framework as a starting point for understanding the user behaviours they will need to achieve in order for their particular model to be successful and realise circular benefits. In addition, the Framework can be used to counter tendencies towards ‘technological optimism’ by supporting the critical evaluation of claims about likelihood of user uptake and acceptance of new business models or circular solutions. It does this through enabling measurement of current adoption levels and experienced barriers and facilitators of adoption, for example. At the same time, the Framework can also be used to prompt the identification of opportunities to improve existing offerings or develop new solutions that enable user circular behaviour.

Future research could also realise the interdisciplinary ‘boundary object’ potential of the Framework in a wide range of ways. One example is by utilising the user behaviour linkages as a departing point to increase connections between research focusing on consumers, and research focusing on producers/businesses. Importantly, understanding the interrelationships also supports ‘sustainability policy as if people mattered’ [75]—tracking the interdependencies and shared responsibilities and connections between powerful actors and publics. It should also be noted that the exercise to identify these relationships was preliminary in nature in that they did not receive the same level of scrutiny from stakeholders as the set of user behaviours. In addition, no similar, previous analysis could be identified to triangulate or validate the results. We therefore propose these outputs as a first iteration of a model of inter-relationships, and both hope for and look forward to future research and discussion about, contributions to and iterations of this analysis by the field.

Broader Implications

Language Matters, and Clarity Is More Important than Consistency

As noted earlier, different groups of stakeholders in different fields, positions and traditions use different language. The Australian stakeholders engaged in this process were no exception. A key finding from our research was that different understanding of language leads to confusion when stakeholders are not aware that others may use language differently. Overturning assumptions of the universality of terms, and recognising that the same words can hold different meanings for different people are critical steps in transdisciplinary collaboration. Once this understanding was reached amongst our stakeholders, it was interesting to note that clarity became prioritised, rather than consistency. The ensuing dialogue suggested that because language choices are driven by many factors including disciplinary background, intended audience and other needs, it is unrealistic to expect adoption of a single, consistent terminology for use in all spheres, particularly in the short-to-medium term.

In light of this, we acknowledge that there is likely language in this Framework that diverges or conflicts with language already in use in various academic fields. In addition, there are almost certainly new labels and terms that have not heretofore been commonly adopted in academic research. It is not our intention per se for the language in this Framework to be adopted by all stakeholders, though we hope it may go some way to building a foundation for consistency. Instead, it is our aspiration for this and future iterations of the Framework to act as ‘Rosetta stone’ [76] to enable those that currently use different terms, or use the same terms differently, to locate their concepts within the Framework in order to communicate with others with clarity.

‘Circular’ Behaviours Do Not Always Support a Circular Economy

A key difference between this Framework and previous attempts is the explicit incorporation of the circular functions or outcomes of the included behaviours. Such an approach is valuable because many of these behaviours do not intrinsically contribute to achieving circular principles. In some cases, from a perspective of ‘net circularity’, these behaviours can actually engender negative environmental impacts [13, 41, 42]. For example, buying second-hand kitchen appliances reduces resource use if it displaces the purchase of new kitchen appliances. If, however, the behaviour is simply used to amass excess appliances, it is not promoting resource efficiency, but rather accumulation [77]. Similarly, donating unwanted clothing can extend the current life of clothing, but when it is undertaken to free up wardrobe space for new purchases, it can actually contribute to the damaging ‘fast fashion’ phenomenon [30]. Camacho-Otero et al. [41] describe this as the need for behaviours to have a ‘sufficiency effect’ in order to realise circular benefits. Including the circular function of each specified behaviour in the Framework ensures that it is clear when, and how, the included behaviours can be considered circular.

Furthermore, the implied hierarchy of behaviours across the sequenced phases is useful for understanding whether certain behaviours are more or less ‘circular’ relative to others. However, more work is needed to substantiate the proposed hierarchy and to illuminate under what conditions the circular benefits of the behaviours are realised. Locating behaviours in sequential, life cycle phases can similarly enable multi-criteria impact analysis which enables consideration of issues of ‘additionality’, spillover and moral budgeting. In this way, the Framework also facilitates linking the ‘social’ and the ‘technical’ systems of circular economy through the lens of behaviour [12, 44]. A key future contribution would be rigorously analysing the ‘material footprint’ [15] potential of different behaviours across the lifecycle, to provide more robust inputs to prioritisation exercises guiding development of interventions.

Some Circular Behaviours Are Not Part of the Formal ‘Economy’

An interesting finding from the comparative analysis of the Framework against existing studies was that the three behaviours incorporated in our framework that were not covered by the existing seven models were all ‘informal’ behaviours; that is, they are not part of any systems of production or exchange that are central to the ‘economy’.

We have made the point consistently that the terms we choose can constrain our thinking, and the choice of the term ‘circular economy’ is no different: the ‘economy’ aspect of circular economy may be limiting understandings of the behaviours and roles required [55], just as our choice of the term ‘user’ did. Others have previously explored the distinction between ‘circular economy’ and ‘circular society’, and the impact the different terms/conceptions/discourses can have [45, 78]. We note that broadening the overarching framing concept to be more inclusive is likely to advance the value of this Framework even further.

Conclusion

Previous literature has clearly identified the need to more deeply consider the role of the consumer/user in a circular economy. An analysis of existing consumer/user research revealed that current efforts are being undertaken without a clear and comprehensive understanding of the specific behaviours that will be required or desired of users in a circular economy.

A transdisciplinary, iterative process was undertaken with over 100 circular economy experts from across Australia’s public, private, civil and research sectors to develop a comprehensive Framework of User Circular Behaviour. The Framework sets out 59 distinct user circular behaviours, clearly defined and categorised according to circular function and user purpose. A comparison of the Framework against existing models of user behaviours reveals a new level of both breadth and depth now available to the field. Furthermore, the Framework provides a starting point for understanding the relationships between User behaviour and that of other actors, particularly the inter-dependency with up- and downstream circular activity. This analysis is critical to understanding the full range of behaviours needed across the value chain in order to successfully realise a transition to the circular economy. We recommend wide adoption of this Framework in research and policy as a key advance in better integrating users, and the diverse stakeholders enabling and constraining their options, as key actors in the transition to a circular economy. This includes further development of the Framework, and we have set out some important avenues for refining and expanding it.