*3.2.2 The consumer behaviour perspective: drivers and barriers*

Two distinguishable research foci can be identified in the literature: Consumer behaviour regarding the sales volume of secondary and primary EEE, as an engine for the success of *e*CLSC [37–39], and consumer *reversing* and out-of-use storage (stockpiling) behaviours (see **Tables 3** and **4**). Recognising this, a recent systematic literature review [35], aimed to synthesise extant findings from the consumer behaviour literature on determinants of purchase, extension of life and EoU/EoL management of EEE. In contrast, our keyword search was restricted to the CLSC and reverse logistics (RL) literature. Consequently, the studies identified hereby focused primarily on the consumer's role regarding WEEE management, such as recycling and reversing behaviours, as shown in **Tables 5** and **6**. Surprisingly, while [35] identify some additional studies focusing on purchasing and lifetime extension behaviours, our search was able to identify further sources regarding disposal and reversing behaviours. This highlights the need for consumer behaviour research on all the phases of consumption of EEE to align etymologically with research on CLSC, RL and CE, to explicitly account for the importance of all the roles of consumers in relation to *e*CLSC. This result indicates that consumer behaviour research on purchase and lifetime extension of EEE is not fully aware of its importance to the context of *e*CLSC.

Despite these differences regarding the sample of articles identified, our findings in terms of theoretical frameworks and behavioural predictors are in coherence with [35]. In particular, as we depict in **Tables 5** and **6**, the identified literature is dominated by theoretical frameworks that are built around the Theory of Planned Behaviour (TPB). While the results support the TPB's utility in understanding the influences of some cognitive factors behind behaviour, they do come with some


#### **Table 5.**

*Target construct and identified predictors for studies exploring no specific WEEE categories. The "Dependent construct" column contains the construct of interest for each study, be it behavioural, attitudinal or otherwise. The "Main predictors" column contains the main predictors of the dependent construct, identified by each study.*


#### **Table 6.**

*Target construct and identified predictors for studies exploring specific WEEE categories explicitly. The "EEE category" column displays the product category addressed by the study. The "Dependent construct" column contains the construct of interest for each study, be it behavioural, attitudinal or otherwise. The "Main predictors" column contains the main predictors of the dependent construct identified by each study.*

limitations. These include the tendency to focus on the formation of intentions under the assumption that these will be strong predictors of actual behaviour, e.g. [42, 46], and the lack of space for the identification of new processes, mechanisms and factors which can be relevant to behavioural outcomes, such as institutional context, personality traits, habitual behaviours or cultural dispositions, among others. The assumption that reported intentions reflect actual behaviour has been identified as an important limitation in the wider sustainable and ethical consumer behaviour literature due to a phenomenon known as the intention-behaviour gap, whereby consumers over-report on their intentions and attitudes when compared to observations of how much they really adopt sustainable-ethical behaviours [49, 50]. Finally, the lack of heterogeneity on theoretical (and methodological) frameworks results in a very good understanding of a very small portion of the plurality of mechanisms, processes and factors that determine consumer behaviour. Therefore, there is a need for theoretical and conceptual innovation regarding consumer behaviour for *e*CLSC, particularly across EEE categories and characteristics as we illustrate in the following sections.

In **Table 5**, we include the articles that do not focus on behaviours regarding no specific EEE category. Conversely, **Table 6** contains the identified studies that explore behaviours focusing on specific EEE categories. In both tables each study is presented with the main construct of interest and the main identified predictors, while in **Table 6**, the EEE categories studied are also specified. A quick inspection of these two tables provides some interesting insights. Specifically, the literature at the intersection between consumer behaviour and *e*CLSC, is dominated by studies on mobile phones specifically (6 studies) and on no specific WEEE category (6 studies). One article was also found to consider ICT equipment more widely [31], another study looked at smartphones more specifically [39], and only one article considered more than one EEE category explicitly [45].

Moreover, in the context of our discussion of different levels of behaviours (see **Tables 3** and **4**), we found most studies to fall within levels 2 and 3 as shown in **Tables 5** and **6**, and only one study considered post-consumption disposal behaviours more generally, but did so by breaking them down into some lower level, less abstract, behaviours [33]. As such, the research literature offers much space for improvement in this sense.

In conclusion, there is a need for innovation regarding the theoretical frameworks employed to enrich academic knowledge on consumer behaviour relevant to *e*CLSC, beyond the constructs hypothesised by TPB and similar frameworks. Moreover, results from studies focusing on the formation of intentions should be interpreted with care due to their susceptibility to biases. There is a need for empirical research to explore consumer behaviours pertaining to a wider variety of behavioural layers and specific EEE categories to establish the fundamental behavioural differences across categories of varying characteristics, the different perspectives that can be achieved through framings of varying abstraction, and the reasons behind them. This is further highlighted below.

#### *3.2.3 EEE characteristics and pre-conditioning of consumer behaviour*

Studies have used consumer surveys to profile the purchasing, usage, stockpiling, replacing and discarding of EEE in different national and sub-national settings (e.g. [51, 52]). For instance, [51] conducted a survey, distributed among a representative sample (n = 395) of households in Sao Paolo, Brazil. The study considers all the stages of consumption and, given the large variety of product categories that fall within the scope of EEE, the authors account and collect data for 26 separate EEE categories. Not surprisingly, their results vary significantly from one EEE category to another, as well as in comparison to similar results from different geographical contexts.

When considering in-use EEE, they found that each household had, on average, 17 items, where mobile phones and cathode-ray-tube TVs (CRT-TV) lead the ranking with about 2 per household. However, a closer look at how long the equipment had been owned, revealed that 60% of the mobile phones had been owned for less than 2 years at the time of reporting, while almost 85% of CRT-TVs had been owned for longer than 2 years, and about 25% had been owned for over 10 years. The latter, however, was true only for a negligible percentage of in-use mobile phones. Similar results are obtained by [44] from a sample of Portuguese consumers, where LCD-TVs are found to be typically in use for more than 10 years, while mobile phones were in use typically for 3-5 years. Given that CRT-TVs were found to have been replaced by LCD-TVs, since only 16% of the former were less than 2 years old while this applied to 74% of LCD-TVs [51], the differences and parallels of these studies likely indicate that CRT-TVs had already been replaced a lot earlier in Portugal. As illustrated by these examples, there are significant differences in consumer purchasing and use behaviour between different EEE categories, which persist across social contexts.

When considering out-of-use EEE that was being stored, i.e. stockpiling, the study finds these to represent about 12% of the total amount of EEE that is present in the surveyed households. The authors attribute this finding to the "treasure effect", a phenomenon whereby consumers tend to overvalue out-of-use products, and consequently do not discard them but instead keep them in a drawer or storage room over the belief that at some point in the future it will be needed and used again. This idea was supported by their findings which reveal that more than 50% of the out-of-use EEE was fully functional for the majority of EEE. However, for washing machines, microwave ovens, electric drills and DVD players, over 50% of the out-of-use EEE is functionally damaged. Therefore, their findings, when it comes to stockpiling behaviour, are significantly different for different EEE categories.

#### *Urban Mining of e-Waste and the Role of Consumers DOI: http://dx.doi.org/10.5772/intechopen.100363*

The authors find similar results concerning the acquisition of EEE from different routes (second hand, new … ), reasons for acquiring the EEE and disposal routes (reuse, recycling … ). Furthermore, [36] conduct a survey in South Korea (n = 2000) where they focus on identifying the current state of adoption of WEEE disposal behaviours and the cognitive factors behind it. The study too finds that products of different characteristics lead to contrasting results and dedicates one section to exploring how their results vary between EEE categories of varying sizes (small, medium and large). Namely, their findings suggest, among other differences, that while take-back initiatives only represent 10.24% of the disposition routes for small and medium sized appliances, it is the leading route for large EEE as it represents 34.5% of the total. However, the authors do not go beyond these characteristics as it is not the focus of their study.

As illustrated by the examples provided above, when considering consumer behaviour, the EEE category under study plays an important role in determining the needs and wants of the consumer, and hence their behaviour. Additionally, very significant differences emerge in similar studies in different national contexts due to important institutional and cultural differences. However, this does not change the fact that there are product characteristics which fundamentally influence consumer perception and behaviour in their decision-making process [45, 53].

Despite that, [36]'s attempt to explicitly assess the behavioural differences among different EEE clusters, namely by size, is one of the only available accounts. Other interesting findings in support of the importance of considering the types of EEE categories include [45]'s article which finds product type to be one of the most critical factors in determining behavioural differences. However, the study fails to unveil what the factors are that differentiate the EEE categories they explore, providing little information on how these EEE characteristics may be operating in leading to different behavioural outcomes. Finally, [30, 31] conduct two studies with the same sample (and dataset) but targeting ICT products in general and mobile phones, respectively (see **Table 4**). Both studies use the TPB to explore the adoption of recycling behaviours for ICT equipment and mobile phones separately. The results change substantially for the behavioural adoption, as intentions are only able to explain about 9% of the variation in the case of mobile phones, in contrast to 15% for ICT in general, while the opposite happens in the case of intentions to recycle, of which about 36% of the variation can be explained for mobile phones, but only 30% is captured in the general case. Additionally, when comparing structural equation modelling results with artificial neural network outcomes, the results agreed in the case of mobile phones, but in the case of general ICT equipment, the two analyses result in differences regarding the predictors'significances relative to one another. This further illustrates the idea that different EEE categories may better fit certain theoretical and methodological frameworks than others.

While all this evidence highlights the importance of developing an understanding of the EEE category characteristics that influence consumer behaviour and the reasons behind these influences, the literature is currently missing a comprehensive framework through which to do so. As [45] conclude, there is a need "[ … ] for the refinement of EEE classifications used for collection operations to encompass consumers' preferences, and not recycling requirements only". Moreover, we further extend their claim to all consumer-behavioural aspects of *e*CLSC, such as acquiring, using and storing, besides the already mentioned disposal.

#### **3.3 A consumer behaviour focused taxonomy of EEE categories**

We begin by stating the intended use of our taxonomy which is the first step in taxonomy building [34]. The intended users of our taxonomy include researchers in the area of WEEE more generally, as well as more specifically consumer researchers in this area. In particular, while there are increasing accounts of behavioural differences among EEE categories, e.g. [51, 54], there are no frameworks that synthesise the main characteristics that may lead to such differences. Our taxonomy aims to provide a foundation on which to build subsequent knowledge regarding the behavioural differences that arise between EEE categories of differing characteristics and that are critical for *e*CLSC success.

Our taxonomy's dimensions and characteristics (see **Section 3.1.2**) are drawn from empirical and conceptual observations in the literature. However, one of the basic goals of a taxonomy is to be easily extendible [34]. In this sense, we provide hereby a starting point on which to further extend our knowledge regarding the main EEE characteristics that lead to behavioural differences among consumers. To construct the taxonomy, we used an initial pool of 29 EEE categories to test the relevance of the taxonomy. The method and the classification of the initial sample of EEE categories is presented in full detail in the *Supplementary Materials* to this chapter, see **Appendix**. The final taxonomy, which we present in **Table 2**, reads as follows:

T4 = {*Size* (Small; Medium; Large; Extra-large); *Involvement* (Low; High); *Long-term reliability expectations* (Low; Moderate; High); *Value type* (Hedonic; Mixed; Functional); *Internet access* (No; Yes); *Multifunctionality* (Low; Moderate; High); *Outdated* (No; Yes); *Social meaning* (Negligible; Moderate; Highly significant)}.

Next, we provide an explanation of each dimension and their characteristics together with some of the potential behavioural differences that may be expected.

#### *3.3.1 Size*

The *Size* dimension contains three characteristics: small (e.g. mobile phones, smartphones, hair drier, electric shaver … ), medium (e.g. coffee machine, DVD player, radio receiver, VCR, laptop … ), large (e.g. LCD-TV, desktop computer, stereo system … ) and extra-large (e.g. washing machine, refrigerator … ). We conceptualise it as having four characteristics since we found the typical *small-mediumlarge* system to be ambiguous when trying to classify our pool of items at the taxonomy building stage. This has interesting implications for consumer research aiming to elicit attitudinal or other differences in consumer perceptions through self-reports. In particular, through disambiguation, grouping EEE categories in four clusters of size, rather than three, could improve discriminant validity of the studies. In other words, while the relative differences between EEE sizes decrease, less of the EEE categories fall within the "boundaries" of the size characteristics, making clustering more natural for respondents of these studies. Size of the EEE is likely to lead to differences in out-of-use storage, use and discarding behaviours [36, 51, 55] of consumers.

#### *3.3.2 Involvement*

The *Involvement* dimension contains two characteristics: low (e.g. batteries, hair drier, electric shaver … ) and high (e.g. washing machine, smart TV, air conditioner … ). This dimension can be broken down into two intertwined aspects: **price** and **risk**. Hence, low-involvement EEE categories fall within a lower price range and have less risk associated with their purchase, while high-involvement ones fulfil the opposite. As the name suggests, these differences in what we call involvement, invoke different levels of interest and importance, quantity and type of information required to reach decisions for the consumption process at hand, in other words this captures how involved the decision making is expected to be. Less involved decision making tends to be dominated by price and routine considerations, while highinvolvement decision making involve premeditation and information seeking leading to a more conscious sequence of decisions.
