Available via license: CC BY 4.0
Content may be subject to copyright.
Citation: Arpin, M.-L.; Leclerc, S.H.;
Lonca, G. The Circular Economy (CE)
Rebound as a Paradox of Knowledge:
Forecasting the Future of the CE–IoT
Nexus through the Global E-Waste
Crisis. Sustainability 2024,16, 6364.
https://doi.org/10.3390/su16156364
Academic Editor: Alan Randall
Received: 15 April 2024
Revised: 23 June 2024
Accepted: 30 June 2024
Published: 25 July 2024
Copyright: © 2024 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
sustainability
Article
The Circular Economy (CE) Rebound as a Paradox of Knowledge:
Forecasting the Future of the CE–IoT Nexus through the Global
E-Waste Crisis
Marie-Luc Arpin 1, *, Stéphanie H. Leclerc 2and Geoffrey Lonca 3
1École de Gestion, Universitéde Sherbrooke, Sherbrooke, QC J1K 2X9, Canada
2
School of Urban Planning, McGill University, Montreal, QC H3A 0C2, Canada; stephanie.h.leclerc@mcgill.ca
3Capgemini Engineering, Technology & Engineering Center (TEC), Industrial Performance, 31703 Blagnac,
France; geoffrey.lonca@capgemini.com
*Correspondence: marie-luc.arpin@usherbrooke.ca
Abstract: There are widespread assumptions to the effect that the real-time data generated through
the 5G-enabled Internet of Things (IoT) will improve material traceability and accelerate the global
transition to a circular economy (CE), thereby helping to achieve the UN Sustainable Development
Goals and carbon neutrality. Many industries, governments, and NGOs are supporting this vision by
investing in related digital infrastructure (5G networks, servers, computer hardware, etc.). Conversely,
recent literature has highlighted a paradoxical phenomenon known as the CE rebound, whereby
sound CE activities end up offsetting environmental gain(s). This challenges the assumption that the
new 5G-enabled IoT will be conducive to greater circularity while carrying its own environmental
weight. Resorting to applied epistemology—a perspective seldom used in sustainability research—
and the global e-waste crisis as an intense case in point, we question the confidence with which actors
predict positive outcomes from the CE–IoT nexus. We argue that avoiding circularity rebounds cannot
be construed as a matter of methodological development or, by extension, modeling sophistication
through real-time data exploitation. Instead, circularity rebounds need to be recognized and theorized
as a paradox of knowledge that also narrows sustainability research’s horizons, despite AND because
of the 5G-enabled IoT. As per this paradox, advanced digital technologies may well be compounding
environmental issues at the same time as they illuminate them.
Keywords: circular economy (CE); 5G-enabled Internet of Things (IoT); CE–IoT nexus; advanced
digital technologies; e-waste; rebound effects; paradox; applied epistemology; intensity sampling
1. Introduction
The circular economy (CE) concept has benefited from widespread public visibility and
political influence in the last decade to the extent that it now stands as a commonly shared
premise of environmental sustainability among government officials, policy designers,
environmental advocacy groups, and businesses. Broadly defined as an approach to
decouple economic growth from natural resource extraction on variable scales [
1
,
2
], CE
can also be construed as a transition pathway in and of itself (e.g., [
3
]) or as an increasingly
popular goal of sustainability transitions research, as it rapidly—if loosely—translates
into political discourse [
4
,
5
]. For the French philosopher and politician Luc Ferry, the
CE concept injects a new momentum into the ecological movement: “If ecology does
not want to be associated with an all-out decline in terms of amenities, comfort, time,
purchasing power, but also freedom and well-being, it will have to stop being punitive,
anti-liberal, diminishing and deadly to become “positive” in finally integrating the notion
of circularity” [6].
Despite such favorable perspectives, the CE concept and its potential outcomes remain
contested [
7
], and its implementation poses multiple social, economic, and technical chal-
lenges [
8
]. This explains, in part, why sustainability experts (professionals and researchers
Sustainability 2024,16, 6364. https://doi.org/10.3390/su16156364 https://www.mdpi.com/journal/sustainability
Sustainability 2024,16, 6364 2 of 22
alike) often tie the future of CE to developments in digital technologies. Indeed, CE’s
operationalization (i.e., the concept’s concrete translation and implementation through poli-
cies, strategic plans, programs, and economic actors’ decision-making) is now increasingly
perceived to rely on an acceleration of widespread digital systems and applications. To-
gether, for example, the 5G-enabled Internet of Things (IoT) and associated devices/sensors
in supply chains [
9
–
11
], edge cloud computing [
11
], Distributed Ledger Technology [
12
],
and/or machine learning for big data analytics [13], are expected to enable and accelerate
traceability and improve the governance of material flows. Whether in line with such
expectations or in tension with them [
8
,
14
,
15
], different kinds of propositions have been
made by researchers to help overcome CE’s operationalization challenges: i.e., propositions
spanning a broad range of disciplinary perspectives in engineering [
2
] and economics [
16
],
but also in sociology [
17
], geography [
18
], or environmental policy studies [
19
]. For a sense
of CE’s conceptual breadth and diversity within and across disciplines, see [7].
In contrast with the disciplinary core of CE research, this conceptual article questions
CE’s operationalization challenges from an applied epistemological perspective. As per
an influential definition by philosopher Larry Laudan: “Applied epistemology in general
is the study of whether systems of investigation that purport to be seeking the truth are
well engineered to lead to true beliefs about the world” [
20
] (p. 2). As such, applied
epistemology “routinely examines [s] truth-seeking practices [
. . .
] to find out whether they
are capable of delivering the goods they seek.” [
20
] (p. 2) Recognizing the newfound
hopes and beliefs of CE’s operationalization brought on by the rise of so-called “intelligent
assets” [
1
]—and perhaps most of all, by the rise of the 5G-enabled IoT and associated
devices/sensors in supply chains, for tracking product and material flows–, we question
whether big data analytics can realistically help fulfill CE’s “green” promises. Treading
in Zink & Geyer (2017)’s footsteps, we specifically ask whether conditions that are both
necessary AND sufficient for the core of CE activities to avoid CE rebounds can ever be
known: i.e., the phenomenon whereby circular systems or activities end up increasing
or accelerating overall production, thereby eliminating partially or fully their expected
environmental benefits [21].
To shed light on this issue, we explore very practical questions about “what [can] be
known or reasonably believed” from the probabilistic system of investigation underpinning
the bulk of CE research and practice [
22
] (p. 153). Our main argument is that full awareness
of these conditions cannot ensue from increasing and optimizing the interface between CE
and advanced digital technologies. As we move on to show, model uncertainty is bound
to increase as modelers tap into the epistemic potential of artificial intelligence (AI) and
big data analytics. No matter how intent modelers may be on improving systems-based
approaches to environmental modeling (e.g., by hybridizing life-cycle assessment using
agent-based modeling, network models, AI techniques
. . .
or a combination thereof) for
sustainability purposes, the exploitation of real-time data provided by 5G-enabled IoTs
exposes the field of sustainability research itself to rebound effects. The question therefore
arises as to whether the knowledge we readily possess about material flows is actionable
without further relying on “intelligent assets”; if it is, how can this knowledge be exploited
to its full potential, despite being “incomplete” or “insufficient” [
23
,
24
]. For example,
on the policy-making side, how might model uncertainty be conceived and exploited as
knowledge instead of being construed as liability, ignorance, or failure [
25
,
26
]? Or, on the
environmental modeling side, how might extant modeling approaches such as Life-Cycle
Assessment (LCA) be epistemically recast so that the modeling community’s core identity
and epistemic values fit more tightly with its modeling practices [
27
,
28
]? Apart from a few
potential exceptions (e.g., [
29
]), such an applied epistemological take on CE has earned no
research attention thus far.
In this article, we set the stage for a new research agenda at the intersection between the
paradox theory in Management & Organization Studies (MOS) and sustainability research.
In Section 2, we offer brief literature and expose our main premises: by adapting Larry
Laudan’s legal epistemology and leveraging it, the CE–IoT nexus is cast as an “epistemic
Sustainability 2024,16, 6364 3 of 22
engine”, urging us to ask whether its mechanics of knowledge production are “truth-
conducive” [
22
]. We develop our main argument in Section 3: i.e., the circularity rebound
phenomenon is theorized as a paradox of knowledge inherent to the CE–IoT nexus. A
conceptual framework based on the paradox theory is also drawn to capture the deadlock
in which the CE’s operationalization is arguably caught. In Section 4, we illustrate the main
argument by drawing from the ongoing electronic waste (e-waste) crisis, framing it as an
intense case in point: that is, as a single case study whereby our phenomenon of interest
(the circularity rebound) is intensely manifested [
30
], and thus exposed as a paradox of
knowledge. Overall, this conceptual article challenges the eco-modernist take on CE’s
operationalization by questioning the effectiveness and enabling capacity of advanced
digital technologies for methodological development and model sophistication.
2. Literature Review: The CE–IoT Nexus as an “Epistemic Engine”
As mentioned above, some of the CE’s most promising operationalization pathways
have been linked to the 5G-enabled IoT: that is, to the epistemic potential of Big Data
and new data analytics in terms of material traceability, monitoring, and control. Tapping
into Larry Laudan’s scholarship in applied epistemology and paradox research in MOS,
we expose the need to unpack some of the key assumptions and shortcomings of this
perspective. Admittedly, our proposition may appear to feed into the false dichotomy
between high- and low-tech operationalization pathways, which currently permeates CE
discourses [
7
,
8
]. To the contrary, however, our premise is that the main challenge lies
elsewhere and much deeper, in our own “trained incapacity” as sustainability researchers
to see beyond data-driven, mathematical modeling and probabilistic analysis for purposes
of sound knowledge production or scientific advice [23,31].
For example, let us consider the long-standing debate on whether and how LCA
modeling can rein in subjective (value-laden) choices by further relying on data and mathe-
matical modeling (e.g., [
32
–
35
]). In line with this last take on the debate, some sustainability
researchers using LCA-based methodologies see value assumptions as a threat to scientific
robustness and credibility. LCA guidelines and ISO standards also implicitly advise model
developers to eliminate values from their models as much as possible [
28
,
35
,
36
]. Yet, as
philosophers of science and social scientists have known and have been arguing for over
three decades [
36
,
37
], “value-based judgments, based on situated knowledge, can actually
enhance the rigor, accountability, and credibility of scientific assessments.” ([
28
], p. 1410).
In becoming more qualified as LCA experts, sustainability researchers are therefore prone
to developing an implicit bias in favor of data-intensive models and metrics, which may ap-
pear more objective and complete than mixed (both qualitative and quantitative) modeling
approaches (e.g., [
38
]): that is, they are likely to refrain from using value-based judgment
even when it would enhance rigor [34].
In this last respect, any scientific training may culminate into what Gendron et al.
coined as “epistemological short-sight” [
23
]: a phenomenon akin to what Hannah Arendt
described as “thoughtlessness” [
39
,
40
]. It follows that sustainability researchers (e.g.,
many of which are engineering or natural sciences scholars) may come to lack theoretical
judgment and/or imagination precisely as they rise and become proficient in their own
discipline and/or field of enquiry [
23
]. Through this process of epistemic acculturation
or “habitus incorporation”, as Pierre Bourdieu would call it [
41
], one typically acquires
“contributory expertise” at the expense of “interactional expertise” [
42
], thus becoming
shortsighted, so to speak, and much less inclined or capable to explore alternate episte-
mologies or take part in interdisciplinary integration [
43
–
45
]. Through this process, CE
research’s prime and longstanding focus on problem-solving becomes exacerbated as if de-
bates about problem definition had no epistemic value of their own or should come without
political considerations [
23
,
37
]. Yet insofar as CE researchers aspire to formulate “socially
robust” responses to the pressing environmental challenges of our time, knowledge and
technology development at the CE–IoT nexus does need to account for such debates and
political considerations [44,45].
Sustainability 2024,16, 6364 4 of 22
Let us, therefore, consider the CE–IoT nexus as the “epistemic engine” it appears
to be in the eye of many sustainability and CE researchers: that is, as a “a tool for
ferreting out the truth from what will often initially be a confusing array of clues and
indicators” [20] (p. 2)
, whereby material traceability is enabled, and the global transition to
a CE made
possible [46,47]
. Let us propose a thought experiment rooted in the legal system
as a metaphor for the CE–IoT nexus, instead of a more familiar form of analysis, to think
about uncertainty, error, and validity in environmental models, the status of knowledge in
environmental decision-making and the operationalization of the CE.
While the 5G-enabled IoT, Big Data and new data analytics are often construed by
sustainability researchers as an epistemic boon, a more critical literature suggests that they
should also be seen as an epistemic bane on at least two fronts:
(1)
In terms of error reduction, i.e., avoiding counterproductive decisions (or “false ver-
dicts”) [
20
], data-driven sustainability science can increase the risk of spurious corre-
lations and may, as such, threaten the very purpose of the CE–IoT nexus [48–50];
(2)
In terms of error distribution, i.e., choosing the lesser evil when errors do occur (e.g.,
supporting “false acquittal” instead of “false convictions”) [
20
], data-driven sustain-
ability science finds itself in the awkward position of being both judge and jury as
it tries to account for the environmental burden of its own sensing-and-measuring
apparatus, namely at the CE–IoT nexus [51,52].
These two sets of issues are not new. Yet, as we argue in Section 3, they tend to be
trivialized, and their underlying paradox is overlooked. As per the paradox definition,
error reduction and error reduction are to be construed as “persistent contradictions between
interdependent elements” [
53
] (p. 10). Such contradictions are persistent in the sense that
they resist any sort of resolution: i.e., paradoxes are inherently insoluble. This is why
aparadox cannot be understood as (or assimilated to) a problem, albeit a very complex
or “wicked” one [
54
]. Because of this intractability, a paradoxical situation needs to be
recognized and managed differently than a problematic one [
54
,
55
]. In fact, a paradox
can become even more damaging when it remains implicit—that is, when it is blindly
assimilated to a problem as if it was plainly amenable to resolution (e.g., [
56
]). So, rather
than considering the CE–IoT nexus in terms of problem-solving, we propose to consider
the dynamics underpinning its mode of knowledge production as paradoxical. Hence, new
insight was gained from this study.
On the first front (error reduction), one must heed the fact that data-driven (as opposed
to theory-driven) discovery increases the risk of assigning meaning to spurious correla-
tions [
49
] and, as such, turns the CE–IoT nexus into both an epistemic boon and bane, all
at once. As statisticians Xiao-Li Meng puts it, Big Data can hold statistical paradises for
the heedful analyst, but it mostly holds fallacies for all those who carelessly turn to it for
scientific breakthroughs based on the false premise that more data necessarily translates
into less uncertainty, better models and better decisions [
57
]. More to the point, “without
taking data quality into account, population inferences with Big Data are subject to a Big
Data Paradox: the more the data, the surer we fool ourselves” [
57
] (p. 686). The same holds
even more true for predictive analysis based on Big Data, whereby “[spurious correlations]
appear only due to the size, not the nature [or quality], of data”, [
49
] (p. 595). It appears that
the future relevance of data-driven sustainability science—and hence the CE–IoT nexus—is
closely tied to the future epistemological proficiency of sustainability researchers [23].
On the second front (error distribution), one must heed the paradox underlying the dig-
ital transformation of environmental modeling, so to speak (e.g., [
58
–
60
]). Epitomizing this
trend, a strand of CE research is reasserting the centrality of methodological development
based on advanced digital technologies to bridge knowledge gaps, reduce uncertainty, and
push back epistemic frontiers. From this perspective, the time is ripe for a major leap in our
collective capacity to monitor and govern material flows in an environmentally optimal
way [
59
]. Thanks to the advent of fifth and sixth-generation (5G, 6G) wireless networks, the
latency and capacity to connect devices worldwide are tremendously improved, thereby
allowing the IoT to become a critical enabler of CE [
61
]. And thanks to recent advances
Sustainability 2024,16, 6364 5 of 22
in lithographic technology and semiconductor traceability, supply chains could now be
tracked and managed down to the wafer level, as opposed to the packaged semiconductor
unit/product (Micro-Electro-Mechanical systems) level [
62
]. A 5G-enabled IoT is therefore
seen as a pragmatic way of fulfilling a long-awaited epistemic function by allowing digi-
tized products, digital components and material flows to become fully traceable and the
data collected to be transformed into real-time, global lifecycle inventories: in other words,
allowing for high-stake environmental decisions and CE policies to be “futureproofed”
based on LCA [
61
]. Yet an oft-silenced precondition to optimizing the CE–IoT nexus is that
the digital infrastructure (5G wireless networks, servers, computer hardware, etc.) itself be
sustainable, or at least, that it be energy efficient [63].
As we move on to show in Section 3, this last precondition may not be attainable,
namely because of “rebound effects” and the myriad uncertainties related to the infor-
mation’s accessibility, how it may be used, or whether it may help inform environmental
decisions and public policies [
21
,
64
,
65
]; in spite, that is, of the great epistemic expectations
the CE–IoT nexus may fuel [66]. As the sensitivity analysis scholar Andrea Saltelli puts it:
There has been an accumulation and maturation of structural contradictions in
modern science [
36
], [
. . .
] between real and acknowledged uncertainty in science’s
pronouncements, between technological progress and technological risk, and in
its purported structural relation to democracy [
67
]. [
. . .
] A radically new concept,
practice, and ethos need[s] to be imagined and acted, by scientists—who need to
be clearer about what they can deliver and what they cannot –, and by society—
which must come to accept a more circumspect understanding of the role of
science in informing societal and technological directions. [68] (p. 87–88)
Saltelli proposes to manage epistemic expectations more responsibly—by being clearer
on “what [scientific models] can [and cannot] deliver”. But what if acknowledging the
extent of inevitable uncertainty ends up reducing confidence in scientists and negatively
impacts public acceptance of scientific predictions and/or environmental advice? [69,70].
As paradox scholar Marianne W. Lewis clarifies, we might be faced with a perception
challenge rather than a matter of clarity and fact: facing paradoxical tensions such as the
ones exposed above, e.g., between scientific knowledge and uncertainty, one tends to frame
the elements in contradiction as conflicting, whereas they are “two sides of the same coin”
whose relatedness tends to be obscured by actors’ perceptions [
71
] (p. 761). Said differently:
“Paradoxes stare us in the face—taunting our established certainties while tempting our
untapped creativity. [
. . .
] While seemingly distinct and oppositional, these elements inform
and define one another, tied in a web of mutuality.” [
55
] (p. 6). This paradoxical perspective
on environmental modeling (and what the latter can and cannot deliver) is reflected in the
sustainability research literature from a philosophy of science standpoint by authors who
conceive of “uncertainty as knowledge.” (e.g., [
25
,
26
]). As [
26
] argues, “uncertainty can be a
source of actionable knowledge rather than an indicator of ignorance” in that, even though
our modern-scientific ability to “look at the past to make predictions about the future [
. . .
]
is now under threat”, this “growing uncertainty about the future [
. . .
] ironically imbues us
with the knowledge of what we can do to escape that uncertain future.” [26] (p. 2–3).
Given the above, the originality of our proposition lies in our framing of the CE
rebound as a “paradox of knowledge” inherent to the CE–IoT nexus and closely akin
to one of the paradoxes currently faced by the LCA community [
27
]: that is, a paradox
whereby uncertainty and knowledge come forth as inherently interdependent elements
bound into a profound and persistent contradiction [
53
,
72
]. This article’s epistemological
rooting, therefore circumvents the barren opposition between techno-optimism and techno-
pessimism, or between science and philosophy, by arguing that the most fundamental
challenge posed by the expansion of the 5G-enabled CE–IoT nexus is that of facing its
underlying paradox: i.e., the CE rebound we define and delve into in the next section
should not be construed as a soluble problem, as “it cannot be solved in the sense of being
tackled effectively, once and for all” [
55
] (p. 15). Once acknowledged as a paradoxical
dynamic, the CE rebound can “be approached as a process which, although at times [under
Sustainability 2024,16, 6364 6 of 22
some circumstances] may appear to be ‘solved’, eventually resurfaces, bringing the action
back to ‘where it all started’—even though every new start is more than a repetition, a
necessarily different start.” [55] (p. 15).
Let us therefore clarify how the CE rebound can be understood as a paradox of
knowledge inherent to the CE–IoT nexus, as opposed to a problem that may eventually be
resolved through the 5G-enabled IoT and related digital infrastructure.
3. Main Theoretical Argument: The CE Rebound as a Paradox of Knowledge
In the fields of operations management [
73
–
78
], sustainability sciences [
79
], and circularity-
related work [
1
], it is widely expected that the digital transformation of industry—more
commonly known as Industry 4.0—will enhance CE and overall environmental sustainabil-
ity. This belief stems from the idea that digital technologies hold an untapped epistemic
potential for integrating efficiency gain and cost reduction in new ways. For example,
technologies underlying the 5G-enabled IoT allow for data on consumer behaviors to be
collected so that material flows can be traced and recovered after the use phase. With
cloud manufacturing, underutilized manufacturing capability can be sensed and taped
into by users from across the world [
76
]. Based on data and AI-driven analytics, organi-
zations could theoretically derive knowledge to design products and services for better
environmental performance over the use phase and through end-of-life (EoL), while at the
same time increasing customer satisfaction [
77
]. Even beyond such opportunities, Industry
4.0 and enabling network technologies appear to offer a multitude of possibilities for the
development of material- and energy-efficient production and consumption systems and,
in turn, for supporting industrial transitions towards the CE (e.g., [
78
,
79
]). The current
academic literature is thus replete with new research projects looking to recognize and
unlock the potential at the CE–IoT nexus, broadly understood (cf. [
79
,
80
]). Yet whether
such CE–IoT solutions can carry their own environmental weight remains debatable [81].
In relation to the above point, two important insights must be considered. On the one
hand, the complexity of the CE rebound has now been aptly conceptualized [
16
,
21
] and
documented (e.g., [
2
,
82
–
84
]). As it stands, the CE rebound can be construed as a frontier
of knowledge in several fields, such as environmental economics, industrial ecology, or
sustainability science in general. That is, the full knowledge that would be necessary to
ensure that CE delivers on its “green” potential—regardless of political will or business
commitment—seems out of reach with respect to the standards of science within these
fields [
21
,
64
]. This is indeed a major reason why part of CE research has recently turned to
so-called “intelligent assets”: i.e., pushing back this epistemic frontier would require the
total traceability that the IoT can presumably deliver [1,85].
On the other hand, however, such “intelligent assets” are creating new and increasing
pressures of their own on the environment, the net balance of which must also be factored
in (i.e., modeled and monitored) to ensure that the purported benefits of CE activities
are not offset by the “intelligent assets” in support of traceability (cf. Section 2,error
distribution). For example, the International Energy Agency (IEA) reported that, in 2021,
only 10% of 83 billion IoT-connected devices and sensors in Europe were generating data
that was being analyzed and put to use [
86
] (p. 17). This over-capacity is on the rise
and can indeed be seen in a positive light as an untapped potential, e.g., for sustainable
policy-making, more traceability, better environmental models, optimized urban planning
and citizen benefits, etc. Yet it comes with its own unforeseeable environmental footprint,
which already requires to be modeled and monitored if CE rebound is to be avoided. In
essence, this is how modeling uncertainty is bound to expand along the same curve as
scientific knowledge: although the most sophisticated systems-based modeling approaches
(e.g., life-cycle assessment, agent-based modeling, deep learning) do reveal the complexity
and interrelatedness of environmental sustainability issues such as economic growth,
climate change or digitization [
64
,
65
], this revealing diversely contributes to expanding
the complexity and opacity of systems [
87
]. As such, methodological development is
arguably bound to fall prey to a paradoxical phenomenon known to management scholars
Sustainability 2024,16, 6364 7 of 22
as “paralysis by analysis”, i.e., the unintended deferral of important actions by aspiring to
comprehensive decision analysis [88–91].
3.1. “Paralysis by Analysis”: Statistical Methods’ Paradoxical Appeal to Sophistication
“In their decision-making activities—as Ann Langley emphasized in 1995–, man-
agers need to tread a fine line between ill-conceived, arbitrary decisions (“extinction by
instinct”) and an unhealthy obsession with numbers, analyses, and reports (“paralysis by
analysis”).” [
88
] (p. 63). Almost three decades later, part of this general idea is strongly
echoed by the newly popular dual-process theories, which are now being applied to a
wide range of decision-making domains, i.e., well beyond their origins in psychology
and behavioral economics [
92
]. As per these theories, decision-makers are now broadly
recognized as prone to cognitive biases, which in effect correspond to the “extinction by in-
stinct” end of [
88
]’s spectrum of decision-making impetuses. Incidentally, by conceiving of
reasoning as an interaction between intuitive (System 1) and deliberate (
System 2)
thought
processes [
93
], dual-process models have catalyzed the design of appealing approaches
to sound decision-making, i.e., approaches that incentivize System 2 reasoning in various
domains (e.g., [
94
–
96
]). As a result, however, the risks of “paralysis by analysis” tend to be
overlooked: the new epistemic attractiveness of fast-rising IoT infrastructure (5G networks,
AI chips, sensors, etc.) attests to this neglect, as such infrastructure tends to be construed
as a means to escape decision biases, and not in the least as a threat to decisiveness. As
with the Big Data Paradox introduced above, we might consider that the more we seek to
advance frontiers of knowledge through analytical models, the more issues reveal them-
selves intractable or “wicked” [
97
] at the CE–IoT nexus. Moreover, as we argue in Section 4
using the global e-waste crisis as a revealing (or intense) case study, the more complexity
we uncover about wicked problems, the more we continue to seek methodological and
technological sophistication to solve them in a headlong and relentless pursuit of analytical
comprehensiveness (e.g., [9,98]).
To illustrate the paradoxical dynamic of knowledge and uncertainty more plainly,
let us imagine the necessity to make a strategic decision in the abstract and let us define
strategic decisions as “important and non-routine decisions that require significant resource
commitments and have notable influences on the performance, survival, and health of
firms.” [
91
] (p. 416). In other words, strategic decisions can be a matter of life and death
for an organization, just like sustainability transitions are a matter of life and death for
humanity, as it were [
99
]. In an article entitled “Method in the madness? A meta-analysis
on the strategic implications of decision comprehensiveness”, Samba et al. [
91
] provide
just what we need to run such a thought experiment: i.e., they find themselves confronted
with “paralysis by analysis” as they delved into the management science literature to ask:
Is strategic decision comprehensiveness beneficial for firms? In answering this question,
they, in fact, chronicle our own paradox of knowledge in a stylized form, thereby setting
the stage for its “real-life” exploration in the following sections.
In their meta-analytical inquiry, Samba et al. [
91
] focus on quantitative studies that
“give sufficient statistical information for collecting and computing correlations” [
91
]
(p. 422)
, in addition to defining a series of content-related requirements for basic qualifica-
tion. After thoroughly vetting the 81 papers retrieved from multiple academic databases,
33 usable
papers spanning 34 years (from 1984 to 2018) are selected to provide the empirical
foundation for the meta-analytic work. Based on this literature, Samba et al. [
91
] intrigu-
ingly conclude that the positive effects of decision comprehensiveness on performance
outcomes are driven by a methodological artifact: i.e., performance is correlated with the
early design choice as to whether a decision’s outcome will be measured (1) through percep-
tual data (scale ratings) generated by key informants in the strategy process, or (2) through
official data drawn by researchers from available archives. They find that comprehensive-
ness and outcomes are positively related only when these outcomes are measured through
perceptual data (labeled as subjective by authors), whereas studies based on official archival
data (labeled as objective) show no correlation between decision comprehensiveness and
Sustainability 2024,16, 6364 8 of 22
outcomes. In other words, the authors conclude that the statistically significant findings of
comprehensiveness’s effects on decision performance (i.e., positive outcomes) likely result
from common method bias and other methodological factors creating fertile ground for
informant biases such as social desirability, anchoring bias and availability heuristic [91].
In light of these findings, Samba et al. [
91
] recommend that future research on deci-
sion comprehensiveness be designed to pinpoint these potential methodological artifacts:
i.e., that future research be “designed to include both objective and subjective outcome
assessments.” [
91
] (p. 431). Yet, in doing so, they tread a fine line between increased
confidence in results through more comprehensive data and diminishing returns from
increased model complexity and ensuing equifinality [
100
]. In other words, they find
themselves confronted with the dangers of “paralysis by analysis”. In the following long
excerpt from [
91
], we add the italic letters in brackets to illustrate how a manager’s and
a researcher’s comprehensiveness ordeal mirrors each other. In spite of having very dif-
ferent takes on performance outcomes than managers, Samba et al. [
91
] find themselves
confronted with the same dilemma. Their conclusions therefore chronicle the paradox of
knowledge we seek to clarify and the symmetry between a manager’s and researcher’s
bias in favor of comprehensiveness:
“Beyond a methodological explanation for our results, we build on the view that
strategic decision [or statistical] comprehensiveness provides symbolic value for
organizations [or research], which can be captured by managers’ [or researchers’]
accounts of organizational performance [i.e., research impact]. [
. . .
] Although ad-
vanced predictive analytics can help managers [and researchers] better predict
complex and uncertain future states, contributors to research on the strategic
implications of big data have a duty to remind readers [and themselves] that “un-
healthy obsession with numbers” and “paralysis by analysis” remain relevant
concerns. [
. . .
] Our findings regarding the comprehensiveness-outcome rela-
tionship suggest that while it provides some benefits, comprehensiveness may
have been oversold. Thus, managers [and researchers] should understand that
comprehensiveness is likely not the sought-after panacea.” [91] (pp. 431–433)
In other words, Samba et al. [
91
] exemplify how the comprehensiveness-outcome
relationship is paradoxical and how it also lurks within research. Most importantly, they
exemplify how statistical methods’ appeal to sophistication endures in spite of one’s
intimate knowledge of its traps.
Now, going back to the CE–IoT nexus, let us show how Samba et al. [
91
]’s own paradox
of knowledge becomes salient to an understanding of the CE rebound. As [
21
] demonstrate,
(1) there is nothing intrinsically green about the circular economy, and
(2) the
methodologi-
cal comprehensiveness sought after to identify necessary and sufficient conditions for CE’s
green character is likely unfeasible. Firstly, “simply closing material loops is not enough
to guarantee environmental improvement” since “increased production and consumption
[
. . .
] can reduce the benefits of the circular economy” [
21
] (p. 600). Such an increase is
what they have defined as CE rebound: i.e., a possible outcome of CE activities whereby an
environmentally sound change in secondary production—i.e., recycling and repair being
less impactful than primary production—fails to substitute primary production (for ex.,
by lowering prices on a given market and prompting greater consumption). Therefore,
if researchers, decision-makers and policy designers were to avoid the CE rebound, they
would have to factor in “the net consequences of increased secondary production” by
“looking at all possible causal chains.” [21] (p. 600).
Secondly, then, in substantiating the CE rebound phenomenon, the authors of [
21
]
raise the bar of robust environmental sustainability and CE research so high that it becomes
unattainable. From an applied epistemology point of view, one might even say that they
reveal the radical incapacity of dominant environmental modeling approaches to “ferret
out the truth” [
20
] (p. 2) from the global economy. Economists Gillingham et al. [
64
]
make a similar argument while discussing the energy efficiency (EE) rebound effect and in
finding the global economy from which it emerges to be a “single, interconnected, complex
Sustainability 2024,16, 6364 9 of 22
dynamic system” whereby “definitive arguments about cause and effect [are made] nearly
impossible.” [64] (p. 81). But as [21] more directly conclude:
Circular economy rebound could be a serious obstacle to creating meaningful
environmental improvement. How, then, can we work to avoid rebound so that
the promise of the circular economy is realized? From the preceding discussion,
several necessary conditions emerge for avoiding circular economy rebound. [
. . .
]
Unfortunately, due to the unpredictable nature of highly complex systems, such as the
system of markets involved in the circular economy, it is likely impossible to derive any
meaningful conditions that are both necessary and sufficient. Unforeseen consequences
may mean that a well-intentioned circular economy activity nonetheless results
in rebound. [21] (p. 599. Emphasis added.)
In other words, the authors of [
21
] expose the impossibility of grasping and preventing
CE rebound(s), or the energy efficiency (EE) rebound(s) for that matter, through advanced
scientific research. They show how “paralysis by analysis” also lurks beneath CE research
and practice, and beneath the sustainability sciences more broadly.
In light of this, some argue—as [
21
] do—that unleashing CE’s “green” potential should
go hand in hand with challenges to economic growth and neoliberal
capitalism [101,102]
such that degrowth scenarios varyingly seem to present themselves as potential responses
to current environmental crises [
103
]. Yet, regardless of CE’s conceptions as strong or weak
forms of sustainability, we argue that the CE rebound substantiated by [
21
] and others after
them provides sufficient evidence of the counterproductive dynamic underlying statistical
inference and predictive analysis from Big Data: i.e., identifying and avoiding the CE
rebound should not be construed mainly as a matter of quantification and/or quantitative
methodological development and, by extension, of technological sophistication [
104
]. In-
stead, the CE rebound needs to be understood as a paradox: i.e., in this case, acknowledged
as the ever-receding frontier of knowledge which it constitutes despite the 5G-enabled
IoT and the new epistemic potential it indeed unlocks. As we clarify in the discussion
(
Section 4.2
), for sustainability researchers and policymakers, this would allow the adop-
tion of a paradoxical mindset [
105
], i.e., “shift[ing] [one’s] expectations from rationality and
linearity to accepting paradoxes [CE rebounds] as persistent.” ([
106
], p. 385). In terms
of CE operationalization, this would imply “a shift from traditional linear management
approaches to more adaptive and integrative strategies” [
107
]; i.e., to more adaptive and
integrative strategies aimed at avoiding the unwitting creation of vicious circles at the CE–
IoT nexus. Let us therefore clarify the paradoxical dynamic unwittingly reinforced by CE
research’s dominant mindset and the unwary promotion of digital technology development
to bridge knowledge gaps and push back epistemic frontiers.
3.2. The CE Rebound as an Ever-Receding Frontier of Knowledge
As seen above, for many sustainability researchers, the time seems to be ripe for a major
leap in our collective capacity to monitor and govern material flows in an environmentally
optimal way [
75
]. Thanks to the advent of fifth-generation (5G) wireless networks, the IoT
can now allegedly become a critical enabler of CE [
61
] by not only allowing for product
lifecycles to become fully traceable [
108
]: that is, for “all possible causal chains” to be looked
at and the net consequences of increased secondary production assessed. But how do we
know if the environmental benefits of exploiting the 5G-enabled IoT for operationalizing
the CE outweigh the related environmental costs [
109
]? More broadly, “in what contexts
can digital technologies mitigate the [CE] rebound effect?” [65].
If the CE rebound is to be avoided through total product traceability and CE markets
transparency, the required 5G network latency and capacity gains must not come at the
cost of an energy efficiency (EE) rebound [
63
]. Yet when it comes to operationalizing and
optimizing the CE through the 5G-enabled IoT, a trade-off between the two rebound effects
(CE and EE) does seem unavoidable: as the following thought experiment shows, the 5G-
enabled CE–IoT nexus which epitomizes the “digital assets” expected to “unlock the circular
economy potential” [
1
] can only seek to avoid CE rebounds by creating or exacerbating EE
Sustainability 2024,16, 6364 10 of 22
rebounds, hence the insight provided by a paradoxical perspective on the CE–IoT nexus,
and the theorization of CE rebounds as an ever-receding frontier of knowledge.
Characterizing this impasse, Sustainable Computing (SC) researchers have been warn-
ing against one-sided rationales and overly optimistic scenarios such as those prevailing
in CE research (e.g., [
85
,
87
]). As [
63
] put it, this is why “energy efficiency, measured in
bits per Joule, has emerged as a new prominent figure of merit and come to be the most
widely adopted green design metric for wireless communication systems.” (p. 72) Despite
all efforts, however, SC researchers foresee the proliferation of EE rebounds triggered by
the 5G networks that are supposed to help neutralize the risks of CE rebounds and unlock
CE’s “green” potential:
With such a tremendously expanding demand for wireless communications in
the future, researchers are currently looking for viable solutions to meet the
stringent throughput requirement. [
. . .
] Even with the [new] 5G paradigms for
improving EE, the power consumption will still grow because of the explosive
data rate requirements in the future. Therefore, improving EE can only alleviate
the power consumption problem to a certain extent and is insufficient for enabling
sustainable 5G communications. [63] (pp. 72–73)
So, what might the options be for those most intensely faced with this last predicament?
For some SC researchers, integrating energy harvesting technologies into future 5G
wireless networks stands out as a new imperative [
63
], whereas for others, the prime solu-
tion seems to lie in the constant energy monitoring and control of 5G-enabled IoT, or ultra-
scale systems (USS) [
109
]: i.e., “high-quality levels of measurement and power-monitoring
infrastructure that can provide timely and accurate feedback to system developers and
application writers so that they can optimize the energy use” [
109
] (p. 40). Regardless of the
chosen option or combination thereof, any new solution layer brings forth its own energy
requirements along with the new flows of monitoring data. In principle, then, “the amount
of monitoring data, its gathering, and processing, should never become a bottleneck nor
profoundly impact the energy efficiency of the overall system” [
109
] (p. 27), yet “similar to
many systems that have to handle large amounts of data, energy monitoring can become
a so-called Big Data system [
. . .
]” [
109
] (p. 40). Moreover still, “the fact that administra-
tors, engineers, and decision makers are increasingly willing to perform predictive and
prescriptive data analyses using interactive interfaces” exacerbates this last scenario [
109
]
(p. 40). As predictive and prescriptive data analyses capitalizing on the 5G-enabled CE–IoT
nexus requires adjusting data collection and monitoring metrics on demand—especially if
CE rebounds are to be anticipated –, even the most sophisticated, leanest SC solutions lead
right into the EE rebound.
Samba et al. [
91
]’s paradox hence resurfaces through SC researchers’ awareness of
their own field’s EE rebounds. Much like the plaintiff in North American criminal courts,
it falls on them to prove the extent of the phenomenon and to prevent it, yet the “system
of investigation” they resort to (i.e., statistical science recast as data science) inevitably
becomes self-incriminating: i.e., the standard of proof “beyond reasonable doubt” thus
turns against itself—so to speak—and becomes impossible to satisfy precisely because
the 5G-enabled IoT constantly improves data monitoring and control. This is how the
CE rebound constitutes an ever-receding frontier of knowledge. Faced with such highly
entangled causal chains and ever more diffuse environmental responsibilities, we might
metaphorically say that SC research is bound by an inoperative standard of proof, whereby
reasonable doubt prevails despite the very high confidence in the “guilt” of the 5G-enabled
CE–IoT nexus, and by extension, of advanced digital infrastructures. This double bind is
familiar ground for climate scientists whose constant struggle has long been to cope with a
concept of scientific modeling fully fit for purpose, but ever unfit to stand the trial of our
most common concept of science [110,111].
So, coming back to the question set forth earlier: what approaches complementary to
statistical inference—or more broadly, to quantitative research—could be used to avoid
“paralysis by analysis”? We argue that intensity sampling holds great potential, i.e., that
Sustainability 2024,16, 6364 11 of 22
of allowing the synthetic knowledge we possess but cannot prove to be true (as per the
prevalent standards of knowledge quality in science) to become actionable. In the following
section, intensity sampling is therefore explained and used. Based on intensity sampling
principles, the e-waste crisis is framed as a paradigmatic illustration of the collective
future awaiting us on many more fronts if progress in sustainability research keeps being
understood primarily by researchers as a matter of methodological development and, by
extension, of scientific modeling and technological sophistication.
4. Discussion: Gaining Perspective through Electronic Waste (E-Waste) Crisis
In contrast with probabilistic sampling approaches, qualitative research methods
based on purposeful sampling allow for intangible, diffuse phenomena to be investigated
while accounting for their intrinsic complexity. Knowledge quality is therefore built as per
very different standards than those commonly shared among the natural sciences: e.g., no
statistical significance is sought. For the sake of discovering greater richness and depth of
insights, theorizing strategies such as the one proposed here prioritize “close data fitting”
over generality or transferability of results [
112
]: that is, much fewer “information-rich cases
whose study will illuminate the questions under study” are selected and studied to “yield
insights and in-depth understanding rather than empirical
generalization” [30] (p. 230)
.
Yet, some level of generalization can be expected and will be recognized varyingly in
what follows.
Our prime purpose is thus to investigate whether and how an intense case of notable
CE failures—i.e., the e-waste crisis—can shed light on the 5G-enabled CE–IoT nexus and its
inherent rebound effect. As we will argue, this case not only provides valuable insight into
the vicious circle created by e-waste management schemes predicated on advanced digital
technologies but also, and more broadly, into the paradoxical dynamics underlying CE’s
operationalization through the CE–IoT nexus. Not only do we acquire more actionable
knowledge from studying this exemplary (information-rich) case than from statistical
depictions [
30
]. But most of all, such an approach contributes to preventing paralysis-
by-analysis by shifting sustainability researchers, policymakers or business managers’
mindset from a traditional, linear problem-solving approach to adaptive and integrative
strategies [
107
]: i.e., a decision-making mindset more cognizant of the paradoxical dynamics
created at the CE–IoT nexus, or whenever sustainability transitions are hinged on advanced
digital technologies.
4.1. Snapshot on the Global E-Waste Crisis
Long before the commercial advent of 5G mobile technologies or the deployment
of IoT edge computing, which are both presented as very promising ways to transition
towards CE [
51
,
113
], production/consumption patterns and EoL management of informa-
tion and communication technologies (ICT) posed critical issues on a global scale, from
an environmental, health, political, social, and humanitarian perspective [
114
–
116
]. For
at least three decades now, these issues have remained intractable despite advances in
modeling capacities, scientific knowledge, targeted legislation, and international agree-
ments (cf. [
117
]). They are intractable to such an extent that the overall situation has come
to be known as the “global e-waste crisis” [
114
,
116
]. In the face of this so-called “wicked
problem” [
97
,
118
], the IoT’s epistemic potential may indeed seem like a promising avenue
to track the life cycles of ICT so as to control e-waste streams in optimal ways (e.g., [
9
,
98
]).
However, notwithstanding the social benefits that the 5G-enabled IoT and advanced data
analytics may present in some circumstances [
119
], these technologies have so far tended
to exacerbate social inequities and environmental harm [119–123].
The challenges posed by e-waste management in many vulnerable contexts represent
an intense illustration of how ICT material flows (and e-waste inflation) cannot be reined
in by resorting to the epistemic potential of ICT itself, through the CE–IoT nexus, artificial
intelligence (AI) and big data analytics. As it unfolds, the global e-waste crisis illuminates
the “dark side” of the CE–IoT nexus as inextricably tied to the epistemic engine it also
Sustainability 2024,16, 6364 12 of 22
constitutes (cf. Section 2): that is, it magnifies the features of CE rebound dynamics, which
arguably span the full spectrum of digital-sophistication-for-the-environment between the
smart city [
118
] and environmental impacts modeling [
124
], thereby making them more
recognizable by sustainability researchers, policymakers, and decision-makers overall. To
support digital-sophistication-for-the-environment, among other uses, layers of ever more pow-
erful wireless networks (LTE, 3G, 4G, and 5G) accumulate, allowing for the proliferation of
various devices, such as individual mobile devices (typically smartphones) and digital data.
Ubiquitous internet access induces a range of new desirable potentialities and epistemic
prospects, but also highly undesirable consequences on human health and ecosystems—the
desirable and undesirable effects being inextricably linked [
125
,
126
]. Even the smallest and
most benign digital devices, such as single IoT edge devices (whose environmental burden
is implicitly assumed to be negligible compared to the positive impacts they can generate),
are very likely to induce important rebounds on a large scale [
51
]. So, although varying in
detail and intensity depending on contexts, the digitization of environmental modeling
or of rapidly urbanizing areas in low-income countries magnifies the trait of an otherwise
diffuse phenomenon whose global social, economic, and environmental ramifications are
still unfathomable: that is, the e-waste crisis.
Almost fifteen years ago, over the course of the year 2008, the number of connected
objects in use on a global scale crossed an important threshold, exceeding the world’s
population for the first time [
127
]. The picture is even more striking given that in this same
year, the “digital divide” was peaking between the Global North and the Global South, both
quantitatively and qualitatively [
128
]. In 2008, it was estimated that less than 3% of Internet
users were in Africa, while the continent held more than 14% of the world’s population [
128
].
Over the next decade, fast-growth urban areas in the Global South (mostly Africa) rapidly
gained connectivity through impressive increases in the number of smartphone users,
but digital inclusion did not follow accordingly [
129
–
131
]. Like the dynamics already
observed in large European or North American cities—i.e., where connectivity is now
nearly universal–, the social (offline) inequities and stratification of digital uses they induce
would largely inhibit the expected benefits of the otherwise essential spread of internet
access [
119
–
121
]. In other words, the internet became increasingly accessible to more and
more people, but this broader access did not translate into lower socioeconomic inequities or
more environmental justice. In some borderline cases, inequities in digital skills and quality
of technological uses even contributed to exacerbating already existing social inequities,
most importantly socio-demographic inequities, despite increased connectivity [119,120].
At the same time, urban areas in many developing countries became the dumping
ground of e-waste for themselves, Europe, and North America [
116
,
120
,
132
]: the fastest-
growing stream of solid waste since the mid-1990s [
133
], and one of the most impactful
locally (on human health, ecosystem integrity, etc.) and globally (on climate change and
critical metal dispersion, etc.) when tied—as it often is—to informal, poorly managed or
unmanaged recycling (see e.g., [
134
]). Most alarmingly, some areas in Africa and Asia
became dumping grounds despite the Basel Convention (e.g., [
117
]) and other elaborate
regulatory frameworks [
135
]: that is, in spite of normative frameworks and sophisticated
governance schemes specifically aimed at reducing or ensuring the control of transboundary
movements of hazardous waste [
136
]. In 2022, “only 22.3 percent (13.8 billion kg) of the
e-waste generated [globally] was documented as properly collected and recycled” [
116
]
(p. 7)
, and nowadays, “even countries with relatively mature e-waste systems have low
collection rates” [
98
]. On a global scale, a considerable share of e-waste is therefore still
illegally exported to “pollution havens” such as in Africa or Asia as a pretense of reuse or
metal recycling [134] (p. 14).
In other words, the numerous Extended Producer Responsibility (EPR) programs in
place for electronic products are still grappling with important attractors of illegal e-waste
trade, such as the high economic value of metals or the general lack of incentives to comply
with formal processing channels [
19
,
98
,
136
]. In response, most EPR programs continue
to build on the presumption that the more data there is, the better it is to inform local
Sustainability 2024,16, 6364 13 of 22
legislative arrangements, economic incentives, or urban design. Yet even today, in many
developed jurisdictions, policymakers and government officials are unable or reluctant
to restrict the recycling of e-waste to those who are known to have the best available
technology, thereby allowing informal e-waste flows to proliferate in competition with
regulated EPR programs [
137
]. Moreover, in some contexts, manufacturers refused until
recently to reveal the most basic information about how much ICT equipment is put on the
market [19,138].
Considering these apparently basic and yet unmet conditions, how can it be expected
that additional IoT-derived data would, by default, improve the situation instead of exacer-
bating the e-waste crisis? Are we not in the presence of a revelatory case of CE rebound
whereby an ever-receding frontier of knowledge becomes salient?
As we argue below, the current deadlock partly stems from the presumption that, given
total e-waste traceability, the knowledge on e-waste flows would become sufficient and that
it would be transferable to decision-makers (public officials, policymakers, business leaders,
etc.): typically in the form of clear conclusions on how to act according to optimal scenarios
(e.g., how to incentivize large scale return of EoL electronics for optimal management,
how to devise optimal Extended Producer Responsibility (EPR) policies, etc.). Yet this
rationale overlooks the paradoxes underlying error reduction and distribution (exposed
in Section 2), whereby the CE–IoT nexus produces uncertainty, so to speak, at the same
time as it produces knowledge: in other words, unless “uncertainty [is recognized as] a
source of actionable knowledge” [
27
], p. 1, data-intensive analytical approaches to e-waste
management are likely to feed into paralysis by analysis, as described in Section 3.1, and
even into the e-waste crisis itself by scaling the demand for, namely, IoT edge devices [
51
].
As the renowned engineer and philosopher Jean-Pierre Dupuy put it back in 2012:
By placing the emphasis on scientific uncertainty [i.e., error reduction and distri-
bution], [we] utterly misconstrue the nature of the obstacle that keeps us from
acting in the face of [collective] catastrophe. The obstacle is not uncertainty, sci-
entific or otherwise; it is the impossibility of believing that the worst is going to
occur. [
. . .
] Even when it is known that it is going to take place, a catastrophe is
not credible: this is the principal obstacle. [139] (pp. 577, 585–586)
That is, it remains unfathomable, especially on a collective scale. As one scientific field
in charge of informing the decisions and actions of policymakers, government officials and
corporate leaders in the face of environmental catastrophe, i.e., of making them believe
that the worst is going to occur, Sustainable Computing research is notably caught in
this double bind, as argued in Section 3.2: i.e., the paradoxical injunction of advancing
our knowledge of the crisis by characterizing and measuring ICT’s contribution to it, for
optimization purposes, while accounting for the environmental burden of its own sensing-
and-measuring apparatus. Hence, the quintessential quality of the field’s relationship to
the CE–IoT nexus, as it specializes in the promotion of an e-waste IoT.
4.2. Neither Hope Nor Hoax: The E-Waste IoT as a Quintessential Paradox of Knowledge
As this snapshot illuminates, e-waste flows (as all other material flows) result from
interactions between highly complex social, economic, political, institutional, and behav-
ioral dynamics [
140
,
141
], all of which vary according to local contexts, globalized markets,
and systemic responses. These dynamics are the result of millions of separate decisions
(explicit or implicit) taken by agents that are (knowingly and not) changing or determining
material trajectories—including sustainability research’s scientific orientations, results, and
advices [
138
]. Whereas enthusiastic promoters of the CE–IoT nexus generally consider that
a cost-effective approach to acquiring Big Data on informal, poorly managed or unmanaged
EoL e-waste flows would allow hotspots identification and a gradual circularization of
the ICT industry (e.g., [
9
,
56
,
69
,
73
,
74
,
98
,
129
], others are more cautious: e.g., Sustainable
Computing researchers [
98
,
109
]. These more cautious promoters of the CE–IoT nexus do
acknowledge that the multiplication of RFID tags, sensors, and other Big Data technologies
complexify the management of e-waste streams [
98
]. Yet contrary to Jean-Pierre Dupuy’s
Sustainability 2024,16, 6364 14 of 22
earlier warning, they cling to the belief that the currently digitizing systems of investigation
purporting to seek true and actionable CE knowledge are well engineered to do so.
As a case in point, let us look at a white paper recently published by the International
Telecommunication Union (ITU), the WEEE Forum, the GSMA and Sofies Group:
The application of digital technologies in the transition to a circular electronics
value chain is still developing and its impact on the space is still not completely
understood or expansively studied. While there is consensus on the benefits
of digitalization in accelerating circularity, it can pose negative impacts as well.
[
. . .
] Such an evolution needs to be complemented with continuing research and
efforts to understand its impact on the electronics value chain and its stakeholders,
considering the most vulnerable such as those working in the informal sector.
Digitalization comes with numerous benefits, but an unregulated space can
alienate some of those involved and potentially underscore inequalities, infringe
on privacy and even create more e-waste. [98] (p. 22, 28)
Using caution, these authors emphasize scientific uncertainty and the need to pursue
analytical comprehensiveness through further research: namely, to understand the impact
(environmental and social) of pairing digital technologies and CE strategies to circularize
the electronics value chain. In other words, they assume that the necessary and sufficient
conditions for an e-waste IoT to achieve its “green” potential can and might even already be
known. As these authors consider, “in general, if IoT devices are taken back and companies
are responsible in their creation, IoT can be more beneficial than detrimental” [
98
] (p. 26).
Yet, the paradox of knowledge underlying the CE–IoT nexus becomes even more salient
here, since such an e-waste IoT would also inevitably have to track and monitor its own
self-induced e-waste (as per the error distribution paradox, described in Section 2).
In other words, the research needed to control the environmental impact of an e-waste
IoT would tread a very fine line between providing the capacity to increase confidence
in decisions or policymaking (through more accessible data and comprehensive deci-
sion analysis—assuming this information is made available to environmental protection
agencies and actors committed to solving the e-waste problem) and its own inevitable
blind-spots and diminishing returns (through modeling approaches whose technological
sophistication also contributes to increasing e-waste). It is likely that, as the economy be-
comes more symbiotic—i.e., as material and energy flows become more intertwined in the
CE–, the expected beneficial outcomes become undermined by increasingly unpredictable
phenomena [
16
,
21
]. This is how the complexity, opacity, and intrinsic unpredictability of
undesirable outcomes such as rebound effects come out as unthinkable for researchers,
policymakers, or business leaders when the CE–IoT nexus is conceived strictly as a solution-
space, i.e., not recognized as conducive to paradoxes. Although we know enough to
support collective action—i.e., we know that rebounds will take place and that they may
well precipitate catastrophe, much like “we know from uncertainty, with near certainty, that
climate change is a problem that must be taken seriously” [
26
] (p. 8)–, these rebounds run
the risk of remaining unbelievable (as in unthinkable, impossible to imagine or accept) in
spite (and because) of our ever-growing capacity to track, monitor and measure them [
139
].
This is perhaps one of the most immediate obstacles to CE’s sound operationalization and
overall environmental sustainability, as anticipated by [139].
Besides, and regardless of inevitable rebounds, there would likely be limited value
to having complete visibility of e-waste flows. Indeed, even if all the conditions were
met to translate this into a 100% recovery rate, it would only contribute a small portion
of the resources needed to meet the current demand for electronics production [
142
] to
which the IoT is contributing. As argued above, one tradeoff is, at minimum, a guarantee
that additional e-waste will be generated but also that particular metals will continue to
be dissipated through the initial extraction and product design, thereby challenging the
recycling of IoT components. The opportunity costs of investing public and private funds
in the CE–IoT nexus may be that other, already “actionable” solutions will be left aside.
Sustainability 2024,16, 6364 15 of 22
As ultimately foregrounded by the idea of circularizing the electronics value chain
through an e-waste IoT, the CE–IoT nexus is, therefore, neither a hoax nor grounds for
hope. Instead, it should be acknowledged as a paradox whereby uncertainty confounds
with knowledge and vice versa. Concretely, this means that CE’s operationalization cannot
be resolved through the CE–IoT nexus. Instead, it is likely that the more we buy into an
ever-persistent race toward traceability and modeling/technological sophistication, the
more environmental issues will become intractable. Despite all the useful digital data that
would admittedly be generated by an e-waste IoT, the conjunction of multiple systemic
transitions required to unleash the knowledge value of this Big Data and be able to use it
to tackle the e-waste crisis is hardly imaginable, let alone implementable, without greater
and yet unmeasurable environmental harm. As previously argued, the rebound effects
brought upon by such systemic transitions are bound to behave as ever-receding frontiers
of knowledge, while any attempt at creating a CE by promoting the IoT is sure to generate
more e-waste. Hence, the intensity value of the e-waste crisis as an information-rich
case whose study from an applied epistemological perspective yields crucial insights: it
illuminates the paradoxical dynamic underlying the CE–IoT nexus and how it has so far
been broadly overlooked by sustainability research.
As explicitly stated in the introduction, our argument is thus 100% theoretical (includ-
ing the case study). As such, it is based on a logical sequence of strongly substantiated
arguments (hence the very high number of bibliographic references, many of which are
also theoretical) coming together to form a thought experiment showing that CE research’s
clinging to quantitative approaches is counterproductive. In that sense, the global e-waste
crisis was selected precisely for its self-sufficient, revelatory quality: i.e., because it strongly
substantiates the argument without having to resort to numerical data—the headlong
pursuit of which—we argued—partly explains why CE research in general, and Sustainable
Computing in particular, is stuck in a vicious circle. Choosing more than one case study
for quantitative comparability reasons would, in fact, have impeded the epistemic move
underlying our main thesis.
So, insofar as the global e-waste crisis is deemed to manifest the aforementioned
knowledge paradox in an intense way, this paper offers two important insights from a
methodological point of view: (1) the paradoxical dynamic underlying the CE–IoT nexus is
indeed widespread, or generalizable to a large extent, albeit mostly under less conspicuous
forms than displayed through the global e-waste crisis; and (2) our interpretations of the
global e-waste crisis altogether avoid distorting or dramatizing the paradox [
30
]. Overall,
then, if we (as sustainability researchers) continue to believe that decision comprehen-
siveness will eventually result from pairing the CE with the IoT, the future of rebound
dynamics is likely to align with the vicious circles manifested intensely through the global
e-waste crisis case study.
5. Conclusions
This paper proposed an original contribution by using an applied epistemological
perspective to clarify the paradoxical nature of the relationship between the CE—which
is widely acknowledged as a premise of environmental sustainability—and the IoT as a
systems infrastructure expected to enable the CE. The paradox of knowledge it exposes at
the root of the CE–IoT nexus challenges the assumption that a 5G-enabled green IoT could
eventually be conducive to greater circularity while carrying its own environmental weight.
Instead, we argue that the confidence with which many industries, governments, and
NGOs are now expecting positive outcomes from the CE–IoT nexus should be questioned.
Most centrally, we question the realism and the political innocuity with which some
sustainability research feeds this expectation by aspiring to capture “all possible causal
chains” underlying a system so complex as the CE.
Contrary to conventional scientific wisdom, however, the reason for this questioning
is not merely that any scientific model is flawed, limited and incomplete, i.e., perfectible
through more research. The reason is that the energy and material flow we aspire to trace,
Sustainability 2024,16, 6364 16 of 22
monitor and control through the CE–IoT nexus cannot be construed as “out there”, i.e.,
as separate from the systems of investigations we use to understand them. The digital
transition of environmental assessment tools and sustainability research surely contributes
to increasing the environmental impact of the phenomenon it is assessing. Yet the extent to
which the CE–IoT nexus exacerbates the phenomenon it is looking to control is bound to
remain in the unknown.
Our perspective on the global e-waste crisis therefore suggests that sustainability
research could achieve greater social service if it diverted itself slightly from the prevalent
goal of seeking analytical comprehensiveness. This historically legitimate endeavor now
seems to be turning on itself, so to speak, through its own critical digital transition. As it
stands, the quest for analytical comprehensiveness may indeed differ from the important
political decisions and actions that sustainability research intends to support in order
to confront the current environmental urgency. Because full awareness of the necessary
and sufficient conditions to avoid CE rebounds cannot be known, striving to discover
these conditions (through new digital technologies or otherwise) may well compound
environmental sustainability issues.
Indeed, the way we disregard paradoxes and their intrinsic trait of persistence leaves
us unable to think and act in the face of present environmental urgency, whereby the line
between knowing and believing fades away. This explains the apparently unreconcilable
perspectives on the CE rebound: on the one hand, promoters of the CE cling to the idea
that “the rebound effect is expected to lessen over time as more goods and services meet
‘circularity’ aspirations and as the economy becomes more service-oriented” [
143
] (p. 191);
and on the other hand, skeptics hold that “the more circular resource flows become, the
more likely it is that it is not waste that is avoided, but rather other eco-efficient uses
of resources” [
16
] (p. 63). However, the main challenge arguably lies elsewhere and
much deeper, in our own “trained incapacity” as researchers to see beyond the self-made
(artificial) oppositions between knowledge and uncertainty, knowledge and belief, and
knowledge and paradox.
On a positive note, the question of how CE research and the broader field of sustain-
ability sciences might integrate, gain from, and extend paradox research by developing
an interface with Management and Organization Studies (MOS) makes for a promising
interdisciplinary research agenda. Although limited in the extent to which it discusses the
literature on proper paradox responses (e.g., through a paradoxical mindset), this paper
constitutes a first move in that direction. For this new research agenda to be furthered, a
key insight gained from this study is that paradoxical dynamics underlying the CE–IoT
nexus are prevalent and that they must be recognized and addressed as such. Henceforth,
uncertainty could become a source of actionable knowledge from which unforeseeable
responses to the environmental crisis might arise. For this new potential to be tapped at
scale, however, uncertainty would have to be more broadly construed and taught as knowl-
edge in and of itself within sustainability research (e.g., [
25
,
26
,
31
,
144
,
145
]): i.e., avoiding
rebound dynamics should not mainly be construed and taught as a matter of quantitative
methodological development or, by extension, of modeling and technological sophisti-
cation. Instead, rebounds should be recognized as paradoxes, i.e., as the ever-receding
frontier of knowledge that they constitute, despite the IoT and the new epistemic potential
unlocked by Big Data analytics. Such a paradoxical (or auto-reflexive) relationship between
research and its own object of inquiry has long been acknowledged in many disciplines in
the social sciences (e.g., [
42
,
44
,
146
,
147
]), but it is still largely omitted in engineering and
the natural sciences, where consequences are nonetheless significant.
Author Contributions: Conceptualization, M.-L.A.; methodology, M.-L.A. and S.H.L.; writing—
original draft preparation, M.-L.A.; writing—review and editing, S.H.L. and G.L. All authors have
read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable for studies not involving humans or animals.
Sustainability 2024,16, 6364 17 of 22
Informed Consent Statement: Not applicable.
Data Availability Statement: Data is contained within the article.
Conflicts of Interest: The authors declare no conflicts of interest.
References
1.
Intelligent Assets: Unlocking the Circular Economy Potential; Ellen MacArthur Foundation: Cowes, UK, 2016. Available online:
https://ellenmacarthurfoundation.org/intelligent-assets-unlocking-the-circular-economy-potential (accessed on 24 April 2023).
2.
Lonca, G.; Muggéo, R.; Imbeault-Tétreault, H.; Bernard, S.; Margni, M. Does material circularity rhyme with environmental
efficiency? Case studies on used tires. J. Clean. Prod. 2018,183, 424–435. [CrossRef]
3.
Luis, P.H. Closing the Loop—An EU Action Plan for Circular Economy; OECD—Organisation for Economic Co-Operation and
Development: Paris, France, 2016. Available online: https://www.eea.europa.eu/policy-documents/com-2015-0614-final
(accessed on 23 July 2024).
4. Audet, R. Transition as discourse. Int. J. Sustain. Dev. 2016,19, 365–382. [CrossRef]
5.
D’Amato, D.; Droste, N.; Allen, B.; Kettunen, M.; Lähtinen, K.; Korhonen, J.; Leskinen, P.; Matthies, B.D.; Toppinen, A. Green,
circular, bio economy: A comparative analysis of sustainability avenues. J. Clean. Prod. 2017,168, 716–734. [CrossRef]
6. Ferry, L. Les Sept Écologies; L’Observatoire: New York, NY, USA, 2021.
7.
Friant, M.C.; Vermeulen, W.J.; Salomone, R. A typology of circular economy discourses: Navigating the diverse visions of a
contested paradigm. Resour. Conserv. Recycl. 2020,161, 104917. [CrossRef]
8. Bauwens, T.; Hekkert, M.; Kirchherr, J. Circular futures: What Will They Look Like? Ecol. Econ. 2020,175, 106703. [CrossRef]
9.
Gu, F.; Ma, B.; Guo, J.; Summers, P.A.; Hall, P. Internet of things and Big Data as potential solutions to the problems in waste
electrical and electronic equipment management: An exploratory study. Waste Manag. 2017,68, 434–448. [CrossRef] [PubMed]
10.
Miaoudakis, A.; Petroulakis, N.; Fysarakis, K.; Alexaki, S.; Alexandirs, G.; Ioannidis, S.; Spanoudakis, G.; Katos, V.; Verikoukis, C.
Pairing a Circular Economy and the 5G-Enabled Internet of Things: Creating a Class of Looping Smart Assets? IEEE Veh. Technol.
Mag. 2020,15, 20–31. [CrossRef]
11.
Akbari, M. Revolutionizing supply chain and circular economy with edge computing: Systematic review, research themes and
future directions. Manag. Decis. 2023, ahead of print. [CrossRef]
12.
Magrini, C.; Nicolas, J.; Berg, H.; Bellini, A.; Paolini, E.; Vincenti, N.; Campadello, L.; Bonoli, A. Using Internet of Things and
Distributed Ledger Technology for Digital Circular Economy Enablement: The Case of Electronic Equipment. Sustainability 2021,
13, 4982. [CrossRef]
13.
Agrawal, R.; Wankhede, V.A.; Kumar, A.; Luthra, S.; Majumdar, A.; Kazancoglu, Y. An Exploratory State-of-the-Art Review of
Artificial Intelligence Applications in Circular Economy using Structural Topic Modeling. Oper. Manag. Res. 2022,15, 609–626.
[CrossRef]
14.
Genovese, A.; Pansera, M. The circular economy at a crossroads: Technocratic eco-modernism or convivial technology for social
revolution? Capital. Nat. Social. 2021,32, 95–113. [CrossRef]
15.
Lowe, B.H.; Genovese, A. What theories of value (could) underpin our circular futures? Ecol. Econ. 2022,195, 107382. [CrossRef]
16. Figge, F.; Thorpe, A.S. The symbiotic rebound effect in the circular economy. Ecol. Econ. 2019,163, 61–69. [CrossRef]
17.
Lazarevic, D.; Valve, H. Narrating expectations for the circular economy: Towards a common and contested European transition.
Energy Res. Soc. Sci. 2017,31, 60–69. [CrossRef]
18.
Desvaux, P. Économie circulaire acritique et condition post-politique: Analyse de la valorisation des déchets en France. Flux 2017,
108, 36. [CrossRef]
19.
Leclerc, S.H.; Badami, M.G. Extended producer responsibility for E-waste management: Policy drivers and challenges. J. Clean.
Prod. 2020,251, 119657. [CrossRef]
20.
Laudan, L. Thinking about Error in the Law. In Social Epistemology: Essential Readings; Oxford University Press: Oxford, UK, 2011;
pp. 271–296.
21. Zink, T.; Geyer, R. Circular economy rebound. J. Ind. Ecol. 2017,21, 593–602. [CrossRef]
22. Coady, D.; Fricker, M. Introduction to Special Issue on Applied Epistemology. J. Appl. Philos. 2017,34, 153–156. [CrossRef]
23.
Gendron, C.; Ivanaj, S.; Girard, B.; Arpin, M.-L. Science-fiction literature as inspiration for social theorizing within sustainability
research. J. Clean. Prod. 2017,164, 1553–1562. [CrossRef]
24.
Mercer, D. Why Popper can’t resolve the debate over global warming: Problems with the uses of philosophy of science in the
media and public framing of the science of global warming. Public Underst. Sci. 2018,27, 139–152. [CrossRef]
25.
Bradshaw, G.A.; Borchers, J.G. Uncertainty as information: Narrowing the science-policy gap. Conserv. Ecol. 2000,4, 7. [CrossRef]
26.
Lewandowsky, S.; Ballard, T.; Pancost, R.D. Uncertainty as knowledge. Philos. Trans. R. Soc. Math. Phys. Eng. Sci. 2015,
373, 20140462. [CrossRef]
27.
Arpin, M.-L.; Merveille, N.; Revéret, J.-P. Between Scientific Orthodoxy and Informal Epistemology: How Is the Tension between
LCA Knowledge and Purpose Resolved in Practice? Ph.D. Thesis, Universitédu Québec àMontréal, Montréal, QC, Canada, 2020.
Available online: http://archipel.uqam.ca/id/eprint/13683 (accessed on 1 January 2024).
Sustainability 2024,16, 6364 18 of 22
28. Freidberg, S. From behind the curtain: Talking about values in LCA. Int. J. Life Cycle Assess. 2018,23, 1410–1414. [CrossRef]
29.
Temesgen, A.; Storsletten, V.; Jakobsen, O. Circular economy–reducing symptoms or radical change? Philos. Manag. 2021,20,
37–56. [CrossRef]
30. Patton, M.; Quinn, Q. Qualitative Research and Evaluation Methods, 3rd ed.; Sage Publications: Thousand Oaks, CA, USA, 2002.
31.
Saltelli, A.; Giampietro, M. What is wrong with evidence based policy, and how can it be improved? Futures 2017,91, 62–71.
[CrossRef]
32.
Hertwich, E.G.; Hammitt, J.K.; Pease, W.S. A Theoretical Foundation for Life-Cycle Assessment: Recognizing the Role of Values
in Environmental Decision Making. J. Ind. Ecol. 2000,4, 13–28. [CrossRef]
33.
Hofstetter, P. The value debate: Ecodesign in a global context are there differences in global values and do they matter? Int. J. Life
Cycle Assess. 2002,7, 62–63. [CrossRef]
34. Nathan, C.; Coles, S. Life Cycle Assessment and Judgement. NanoEthics 2020,14, 271–283. [CrossRef]
35.
Lo Piano, S.; Benini, L. A critical perspective on uncertainty appraisal and sensitivity analysis in life cycle assessment. J. Ind. Ecol.
2022,26, 763–781. [CrossRef]
36.
Ravetz, J.R. Postnormal Science and the maturing of the structural contradictions of modern European science. Futures 2011,43,
142–148. [CrossRef]
37.
Longino, H.E. Science as Social Knowledge: Values and Objectivity in Scientific Inquiry; Princeton University Press: Princeton, NJ,
USA, 1990.
38.
Andersen, I.; Ishii, N.; Brooks, T.; Cummis, C.; Fonseca, G.; Hillers, A.; Macfarlane, N.; Nakicenovic, N.; Moss, K.; Rockström, J.;
et al. Defining ‘science-based targets’. Natl. Sci. Rev. 2021,8, nwaa186. [CrossRef]
39. Arendt, H. Thinking and Moral Considerations: A Lecture. Soc. Res. 1971,38:3, 417–446.
40.
Arendt, H. Hannah Arendt Papers: Speeches and Writings File, -1975; Essays and lectures; Remarks; American Society of
Christian Ethics, Richmond, Va. 1973. [Manuscript/Mixed Material] Retrieved from the Library of Congress. Available online:
https://www.loc.gov/item/mss1105601295/ (accessed on 23 June 2024).
41. Bourdieu, P. The peculiar history of scientific reason. Sociol. Forum 1991,6, 3–26. [CrossRef]
42.
Collins, H.; Evans, R. Rethinking Expertise; Paperback edition; University of Chicago Press: Chicago, IL, USA; London, UK, 2009.
43.
Hoffmann, S.; Weber, C.; Mitchell, C. Principles for Leading, Learning, and Synthesizing in Inter- and Transdisciplinary Research.
BioScience 2022,72, 963–977. [CrossRef]
44.
Callard, F.; Fitzgerald, D. Rethinking Interdisciplinarity across the Social Sciences and Neurosciences; Palgrave Macmillan: London,
UK, 2015.
45.
Hoffmann, S.; Deutsch, L.; Klein, J.T.; O’Rourke, M. Integrate the integrators! A call for establishing academic careers for
integration experts. Humanit. Soc. Sci. Commun. 2022,9, 147. [CrossRef]
46.
Rejeb, A.; Suhaiza, Z.; Rejeb, K.; Seuring, S.; Treiblmaier, H. The Internet of Things and the circular economy: A systematic
literature review and research agenda. J. Clean. Prod. 2022,350, 131439. [CrossRef]
47.
Mboli, J.S.; Thakker, D.; Mishra, J.L. An Internet of Things-enabled decision support system for circular economy business model.
Softw. Pract. Exp. 2022,52, 772–787. [CrossRef]
48. Kitchin, R. Big Data, new epistemologies and paradigm shifts. Big Data Soc. 2014,1, 2053951714528481. [CrossRef]
49. Calude, C.S.; Longo, G. The Deluge of Spurious Correlations in Big Data. Found. Sci. 2017,22, 595–612. [CrossRef]
50. Smith, G. Data mining fool’s gold. J. Inf. Technol. 2020,35, 182–194. [CrossRef]
51.
Pirson, T.; Bol, D. Assessing the embodied carbon footprint of IoT edge devices with a bottom-up life-cycle approach. J. Clean.
Prod. 2021,322, 128966. [CrossRef]
52.
Ligozat, A.-L.; Lefevre, J.; Bugeau, A.; Combaz, J. Unraveling the Hidden Environmental Impacts of AI Solutions for Environment
Life Cycle Assessment of AI Solutions. Sustainability 2022,14, 5172. [CrossRef]
53.
Schad, J.; Lewis, M.W.; Raisch, S.; Smith, W.K. Paradox research in management science: Looking back to move forward. Acad.
Manag. Ann. 2016,10, 5–64. [CrossRef]
54.
Arpin, M.-L. L’Ingénieur: Entre Problème et Paradoxe. Ph.D. Thesis, Universitédu Québec àMontréal, Montréal, QC, Canada,
2020. Available online: http://archipel.uqam.ca/id/eprint/13683 (accessed on 1 January 2024).
55.
Cunha, M.P.; Clegg, S. Persistence in Paradox. In Dualities, Dialectics, and Paradoxes in Organizational Life; Oxford University Press:
Oxford, UK, 2018.
56.
Berti, M.; Simpson, A.V. The dark side of organizational paradoxes: The dynamics of disempowerment. Acad. Manag. Rev. 2021,
46, 252–274. [CrossRef]
57.
Meng, X.-L. Statistical paradises and paradoxes in big data (I): Law of large populations, big data paradox, and the 2016 US
presidential election. Ann. Appl. Stat. 2018,12, 685–726. [CrossRef]
58.
Ghoroghi, A.; Rezgui, Y.; Petri, I.; Beach, T. Advances in application of machine learning to life cycle assessment: A literature
review. Int. J. Life Cycle Assess. 2022,27, 433–456. [CrossRef]
59.
Zhang, A.; Zhong, R.Y.; Farooque, M.; Kang, K.; Venkatesh, V.G. Blockchain-based life cycle assessment: An implementation
framework and system architecture. Resour. Conserv. Recycl. 2020,152, 104512. [CrossRef]
Sustainability 2024,16, 6364 19 of 22
60.
Iglesias-Suarez, F.; Gentine, P.; Solino-Fernandez, B.; Beucler, T.; Pritchard, M.; Runge, J.; Eyring, V. Causally-Informed Deep
Learning to Improve Climate Models and Projections. J. Geophys. Res. Atmos. 2024,129, e2023JD039202. [CrossRef]
61.
Pagoropoulos, A.; Pigosso, D.C.A.; McAloone, T.C. The Emergent Role of Digital Technologies in the Circular Economy: A
Review. Procedia CIRP 2017,64, 19–24. [CrossRef]
62.
Canu, C.; Beaudry, R.; Delafosse, M.J.A. Die-Level Serialization Platform Demonstration. In Proceedings of the 2024 35th Annual
SEMI Advanced Semiconductor Manufacturing Conference (ASMC), Albany, NY, USA, 13 May 2024; pp. 1–5.
63.
Wu, Q.; Li, G.Y.; Chen, W.; Ng, D.W.K.; Schober, R. An overview of sustainable green 5G networks. IEEE Wirel. Commun. 2017,24,
72–80. [CrossRef]
64.
Gillingham, K.; Rapson, D.; Wagner, G. The rebound effect and energy efficiency policy. Rev. Environ. Econ. Policy 2016,10, 1–37.
[CrossRef]
65.
Castro, C.G.; Trevisan, A.H.; Pigosso, D.C.A.; Mascarenhas, J. The rebound effect of circular economy: Definitions, mechanisms
and a research agenda. J. Clean. Prod. 2022,345, 131136. [CrossRef]
66.
Giampietro, M.; Funtowicz, S.O. From elite folk science to the policy legend of the circular economy. Environ. Sci. Policy 2020,109,
64–72. [CrossRef]
67. Saltelli, A.; Funtowicz, S. What is science’s crisis really about? Futures 2017,91, 5–11. [CrossRef]
68. Saltelli, A. Why science’s crisis should not become a political battling ground. Futures 2018,104, 85–90.
69.
Howe, L.C.; MacInnis, B.; Krosnick, J.A.; Markowitz, E.M.; Socolow, R. Acknowledging uncertainty impacts public acceptance of
climate scientists’ predictions. Nat. Clim. Chang. 2019,9, 863–867. [CrossRef]
70.
Freudenburg, W.R.; Gramling, R.; Davidson, D.J. Scientific Certainty Argumentation Methods (SCAMs): Science and the Politics
of Doubt*. Sociol. Inq. 2008,78, 2–38. [CrossRef]
71. Lewis, M.W. Exploring paradox: Toward a more comprehensive guide. Acad. Manag. Rev. 2000,25, 760–776. [CrossRef]
72.
Lewis, M.W.; Smith, W.K. Reflections on the 2021 AMR Decade Award: Navigating Paradox Is Paradoxical. Acad. Manag. Rev.
2022,47, 528–548. [CrossRef]
73.
Felsberger, A.; Reiner, G. Sustainable Industry 4.0 in Production and Operations Management: A Systematic Literature Review.
Sustainability 2020,12, 7982. [CrossRef]
74.
Bordeleau, F.-È.; Santa-Eulalia, L.A.; Mosconi, E. Digital transformation framework: Creating sensing, smart, sustainable and
social (Sˆ4) organisations. In Proceedings of the 54th Hawaii International Conference on System Sciences, HICSS-54, Kauai, HI,
USA, 5 January 2021.
75.
Rajput, S.; Singh, S.P. Industry 4.0 Model for circular economy and cleaner production. J. Clean. Prod. 2020,277, 123853. [CrossRef]
76.
Cheng, T.; Xu, Z. Advanced manufacturing systems: Supply-demand matching of manufacturing resource based on complex
networks and Internet of Things. Enterp. Inf. Syst. 2018,12, 780–797. [CrossRef]
77.
de Sousa Jabbour, A.B.L.; Jabbour, C.J.C.; Foropon, C.; Godinho Filho, M. When titans meet–Can industry 4.0 revolutionize the
environmentally-sustainable manufacturing wave? The role of critical success factors. Technol. Forecast. Soc. Chang. 2018,132,
18–25. [CrossRef]
78.
Hatzivasilis, G.; Christodoulakis, N.; Tzagkarakis, C.; Ioannidis, S.; Demetriou, G.; Fysarakis, K.; Panayiotou, M. The CE-IoT
framework for green ICT organizations: The interplay of CE-IoT as an enabler for green innovation and e-waste management in
ICT. In Proceedings of the 2019 15th International Conference on Distributed Computing in Sensor Systems (DCOSS), Santorini
Island, Greece, 29–31 May 2019; IEEE: Piscataway, NJ, USA, 2019; pp. 436–442.
79.
Nobre, G.C.; Tavares, E. Scientific literature analysis on big data and internet of things applications on circular economy: A
bibliometric study. Scientometr. Int. J. Quant. Asp. Sci. Sci. Commun. Sci. Sci. Policy 2017,111, 463–492. [CrossRef]
80.
Nižeti´c, S.; Šoli´c, P.; López-de-Ipiña González-de-Artaza, D.; Patrono, L. Internet of Things (IoT): Opportunities, issues and
challenges towards a smart and sustainable future. J. Clean. Prod. 2020,274, 122877. [CrossRef] [PubMed]
81. Hittinger, E.; Jaramillo, P. Internet of Things: Energy boon or bane? Science 2019,364, 326–328. [CrossRef] [PubMed]
82.
Tamar, M. David Font Vivanco Does the Circular Economy Grow the Pie? The Case of Rebound Effects from Smartphone Reuse.
Front. Energy Res. 2018,6, 39. [CrossRef]
83.
Warmington-Lundström, J.; Laurenti, R. Reviewing circular economy rebound effects: The case of online peer-to-peer boat
sharing. Resour. Conserv. Recycl. X 2020,5, 100028. [CrossRef]
84.
Siderius, T.; Poldner, K. Reconsidering the circular economy rebound effect: Propositions from a case study of the Dutch Circular
Textile Valley. J. Clean. Prod. 2021,293, 125996. [CrossRef]
85.
Ramos, G.; Müller, J.M.; Lees, M. Introduction to the OECD-IIASA Strategic Partnership: The Potential of Systems Analysis
for Addressing Global Policy Challenges in the 21st Century. In Systemic Thinking for Policy Making; OECD Publishing: Paris,
France, 2020.
86.
IEA Secreteriat. Empowering Cities for a Net Zero Future: Unlocking Resilient, Smart, Sustainable Urban Energy Systems; International
Energy Agency (IEA): Paris, France, 2021. Available online: https://iea.blob.core.windows.net/assets/4d5c939d-9c37-490b-bb53-
2c0d23f2cf3d/G20EmpoweringCitiesforaNetZeroFuture.pdf (accessed on 23 July 2024).
87.
Burrell, J. How the machine ‘thinks’: Understanding opacity in machine learning algorithms. Big Data Soc. 2016,3,
2053951715622512. [CrossRef]
88. Langley, A. Between’paralysis by analysis’ and’extinction by instinct’. MIT Sloan Manag. Rev. 1995,36, 63.
Sustainability 2024,16, 6364 20 of 22
89.
Eisenhardt, K.M. Speed and Strategic Choice: How Managers Accelerate Decision Making. Calif. Manag. Rev. 1990,32, 39–54.
[CrossRef]
90.
Cabantous, L.; Gond, J.-P. Rational Decision Making as Performative Praxis: Explaining Rationality’s Éternel Retour. Organ. Sci.
2011,22, 573–586. [CrossRef]
91.
Samba, C.; Tabesh, P.; Thanos, I.C.; Papadakis, V.M. Method in the madness? A meta-analysis on the strategic implications of
decision comprehensiveness. Strateg. Organ. 2021,19, 414–440. [CrossRef]
92.
De Neys, W.; Pennycook, G. Logic, Fast and Slow: Advances in Dual-Process Theorizing. Curr. Dir. Psychol. Sci. 2019,28, 503–509.
[CrossRef]
93. Kahneman, D. Thinking, Fast and Slow; Macmillan: New York, NY, USA, 2011.
94.
Calabretta, G.; Gemser, G.; Wijnberg, N.M. The Interplay between Intuition and Rationality in Strategic Decision Making: A
Paradox Perspective. Organ. Stud. 2017,38, 365–401. [CrossRef]
95.
Rousseau, D.M. The Realist Rationality of Evidence-Based Management. Acad. Manag. Learn. Educ. 2020,19, 415–424. [CrossRef]
96.
Dyer, J.S.; Smith, J.E. Innovations in the Science and Practice of Decision Analysis: The Role of Management Science. Manag. Sci.
2021,67, 5364–5378. [CrossRef]
97. Rittel, H.W.; Webber, M.M. Dilemmas in a general theory of planning. Policy Sci. 1973,4, 155–169. [CrossRef]
98.
Sherpa, Y.; Sinha, D. Digital Solutions for an Electronics Value Chain; International Telecommunication Union, the WEEE Forum, the
GSMA and Sofies Group: Geneva, Switzerland, 2021; p. 39.
99.
Rockstrom, J.; Steffen, W.; Noone, K.; Persson, A.; Lambin, E.F.; Lenton, T.M.; Scheffer, M.; Folke, C.; Schellnhuber, H.J.; Nykvist,
B.; et al. A safe operating space for humanity: Identifying and quantifying planetary boundaries that must not be transgressed
could help prevent human activities from causing unacceptable environmental change, argue Johan Rockstrom and colleagues.
Nature 2009,461, 472. [CrossRef] [PubMed]
100.
Byrne, P.; McAllister, P.; Wyatt, P. Precisely wrong or roughly right? An evaluation of development viability appraisal modelling.
J. Financ. Manag. Prop. Constr. 2011,16, 249–271. [CrossRef]
101.
Easterlin, R.A. Does economic growth improve the human lot? Some empirical evidence. In Nations and Households in Economic
Growth; Elsevier: Amsterdam, The Netherlands, 1974; pp. 89–125.
102. Boltanski, L.; Chiapello, È.Le Nouvel Esprit du Capitalisme; NRF Essais; Gallimard: Paris, France, 1999.
103.
Kerschner, C.; Wächter, P.; Nierling, L.; Ehlers, M.-H. Degrowth and Technology: Towards feasible, viable, appropriate and
convivial imaginaries. J. Clean. Prod. 2018,197, 1619–1636. [CrossRef]
104. Saltelli, A. Ethics of quantification or quantification of ethics? Futures 2020,116, 102509.
105.
Schad, J.; Bansal, P. Seeing the forest and the trees: How a systems perspective informs paradox research. J. Manag. Stud. 2018,55,
1490–1506. [CrossRef]
106.
Smith, W.K.; Lewis, M.W. Toward a theory of paradox: A dynamic equilibrium model of organizing. Acad. Manag. Rev. 2011,36,
381–403.
107.
Schulte, M.N.; Paris, C.M. Working the system—An empirical analysis of the relationship between systems thinking, paradoxical
cognition, and sustainability practices. Corp. Soc. Responsib. Environ. Manag. 2024; early view. [CrossRef]
108.
Gligoric, N.; Krco, S.; Hakola, L.; Vehmas, K.; De, S.; Moessner, K.; Jansson, K.; Polenz, I.; Van Kranenburg, R. Smarttags: IoT
product passport for circular economy based on printed sensors and unique item-level identifiers. Sensors 2019,19, 586. [CrossRef]
[PubMed]
109.
Almeida, F.; Assunção, M.D.; Barbosa, J.; Blanco, V.; Brandic, I.; Da Costa, G.; Dolz, M.F.; Elster, A.C.; Jarus, M.; Karatza, H.D.;
et al. Energy monitoring as an essential building block towards sustainable ultrascale systems. Sustain. Comput. Inform. Syst.
2018,17, 27–42. [CrossRef]
110.
Maibach, E.; Leiserowitz, A.; Cobb, S.; Shank, M.; Cobb, K.M.; Gulledge, J. The legacy of climategate: Undermining or revitalizing
climate science and policy? Wiley Interdiscip. Rev. Clim. Chang. 2012,3, 289–295.
111.
Raman, S.; Pearce, W. Learning the lessons of Climategate: A cosmopolitan moment in the public life of climate science. Wiley
Interdiscip. Rev. Clim. Chang. 2020,11, e672.
112. Langley, A. Strategies for theorizing from process data. Acad. Manag. Rev. 1999,24, 691–710.
113.
Hui, H.; Ding, Y.; Shi, Q.; Li, F.; Song, Y.; Yan, J. 5G network-based Internet of Things for demand response in smart grid: A
survey on application potential. Appl. Energy 2020,257, 113972. [CrossRef]
114.
Widmer, R.; Oswald-Krapf, H.; Sinha-Khetriwal, D.; Schnellmann, M.; Böni, H. Global perspectives on e-waste. Environ. Impact
Assess. Rev. 2005,25, 436–458. [CrossRef]
115.
Pickren, G. Political Ecologies of Electronic Waste: Uncertainty and Legitimacy in the Governance of E-Waste Geographies.
Environ. Plan. Econ. Space 2014,46, 26–45. [CrossRef]
116.
Baldé, C.P.; Kuehr, R.; Yamamoto, T.; McDonald, R.; D’Angelo, E.; Althaf, S.; Bel, G.; Deubzer, O.; Fernandez-Cubillo,
E.; Forti, V.; et al. International Telecommunication Union (ITU) and United Nations Institute for Training and Research
(UNITAR). 2024. Global E-waste Monitor 2024. Geneva/Bonn. Pdf Version: 978-92-61-38781-5. 2024. Available online:
https://www.unitar.org/about/news-stories/press/global-e-waste-monitor-2024-electronic-waste-rising-five-times-faster-
documented-e-waste-recycling (accessed on 24 April 2024).
Sustainability 2024,16, 6364 21 of 22
117.
United Nations Environment Program (UNEP). Basel Convention on the Control of Transboundary Movements of Hazardous Wastes and
Their Disposal; United Nations Environment Program (UNEP): Gigiri Nairobi, Kenya, 1989. Available online: https://www.basel.
int/Portals/4/download.aspx?d=UNEP-CHW-IMPL-CONVTEXT.English.pdf (accessed on 23 July 2024).
118. Colding, J.; Barthel, S.; Sörqvist, P. Wicked Problems of Smart Cities. Smart Cities 2019,2, 512–521. [CrossRef]
119. Taddeo, M.; Floridi, L. How AI can be a force for good. Science 2018,361, 751–752. [CrossRef]
120.
Clark, L.A.; Clark, W.R. Global E-waste: Unintended Consequences of Marketing Strategies Necessitates A Plan for Change. J.
Manag. Issues 2019,31, 331–347.
121.
Bhaskar, K.; Kumar, B. Electronic waste management and sustainable development goals. J. Indian Bus. Res. 2019,11, 120–137.
[CrossRef]
122.
Vinuesa, R.; Azizpour, H.; Leite, I.; Balaam, M.; Dignum, V.; Domisch, S.; Felländer, A.; Langhans, S.D.; Tegmark, M.; Fuso Nerini,
F. The role of artificial intelligence in achieving the Sustainable Development Goals. Nat. Commun. 2020,11, 1–10. [CrossRef]
[PubMed]
123.
Bender, E.M.; Gebru, T.; McMillan-Major, A.; Shmitchell, S. On the Dangers of Stochastic Parrots Can Language Models Be Too
Big? In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, Virtual Event, 3–10 March 2021;
pp. 610–623.
124.
Oduque de Jesus, J.; Oliveira-Esquerre, K.; Lima Medeiros, D. Integration of Artificial Intelligence and Life Cycle Assessment
Methods. IOP Conf. Ser. Mater. Sci. Eng. 2021,1196, 012028. [CrossRef]
125.
Matthess, M.; Kunkel, S. Structural change and digitalization in developing countries: Conceptually linking the two transforma-
tions. Technol. Soc. 2020,63, 101428. [CrossRef]
126.
Santarius, T.; Pohl, J.; Lange, S. Digitalization and the decoupling debate: Can ICT help to reduce environmental impacts while
the economy keeps growing? Sustainability 2020,12, 7496. [CrossRef]
127. Evans, O. Digital politics: Internet and democracy in Africa. J. Econ. Stud. 2019,46, 169–191. [CrossRef]
128. Fuchs, C.; Horak, E. Africa and the digital divide. Telemat. Inform. 2008,25, 99–116. [CrossRef]
129.
De Lannoy, A. Youth, Deprivation and the Internet in Africa; Policy Paper 4, Series 5; RIA—Research ICT Africa: Cape Town, South
Africa, 2018.
130.
Scheerder, A.; Van Deursen, A.; Van Dijk, J. Determinants of Internet skills, uses and outcomes. A systematic review of the
second-and third-level digital divide. Telemat. Inform. 2017,34, 1607–1624. [CrossRef]
131.
Robinson, L.; Cotten, S.R.; Ono, H.; Quan-Haase, A.; Mesch, G.; Chen, W.; Schulz, J.; Hale, T.M.; Stern, M.J. Digital inequalities
and why they matter. Inf. Commun. Soc. 2015,18, 569–582. [CrossRef]
132.
Asante, K.A.; Amoyaw-Osei, Y.; Agusa, T. E-waste recycling in Africa: Risks and opportunities. Curr. Opin. Green Sustain. Chem.
2019,18, 109–117. [CrossRef]
133. Grant, R.; Oteng-Ababio, M. Mapping the invisible and real “African” economy: Urban e-waste circuitry. Urban Geogr. 2012,33,
1–21. [CrossRef]
134.
Forti, V.; Balde, C.P.; Kuehr, R.; Bel, G. The Global E-Waste Monitor 2020: Quantities, Flows and the Circular Economy Potential;
United Nations University/United Nations Institute for Training and Research, International Telecommunication Union, and
International Solid Waste Association: Bonn, Germany; Geneva, Switzerland; Rotterdam, The Netherlands, 2020.
135.
Lifset, R.; Atasu, A.; Tojo, N. Extended producer responsibility: National, international, and practical perspectives. J. Ind. Ecol.
2013,17, 162–166. [CrossRef]
136.
Thapa, K.; Vermeulen, W.J.V.; Deutz, P.; Olayide, O. Ultimate producer responsibility for e-waste management–A proposal for just
transition in the circular economy based on the case of used European electronic equipment exported to Nigeria. Bus. Strategy
Dev. 2023,6, 33–52. [CrossRef]
137.
WEEE Forum. Legally Binding Quality Standards for WEEE Treatment—Action Needed! Available online: https://weee-forum.
org/ws_news/news-2/ (accessed on 15 November 2021).
138.
Leclerc, S.H.; Badami, M.G. Informal E-Waste Flows in Montréal: Implications for Extended Producer Responsibility and
Circularity. Environ. Manag. 2023,72, 1032–1049. [CrossRef] [PubMed]
139. Dupuy, J.-P. The Precautionary Principle and Enlightened Doomsaying. Rev. Métaphys. Morale 2012,76, 577. [CrossRef]
140.
Binder, A. For Love and Money: Organizations’ Creative Responses to Multiple Environmental Logics. Theory Soc. 2007,36,
547–571. [CrossRef]
141.
Brunner, P.H.; Rechberger, H. Handbook of Material Flow Analysis: For Environmental, Resource, and Waste Engineers, 2nd ed.; 1
Online Resource; CRC Press: Boca Raton, FL, USA, 2016; 454p. Available online: https://public.ebookcentral.proquest.com/
choice/publicfullrecord.aspx?p=4771740 (accessed on 23 July 2024).
142.
Elshkaki, A.; Graedel, T.E.; Ciacci, L.; Reck, B.K. Resource Demand Scenarios for the Major Metals. Environ. Sci. Technol. 2018,52,
2491–2497. [CrossRef] [PubMed]
143.
Schröder, P.; Bengtsson, M.; Cohen, M.; Dewick, P.; Hofstetter, J.; Sarkis, J. Degrowth within–Aligning circular economy and
strong sustainability narratives. Resour. Conserv. Recycl. 2019,146, 190–191. [CrossRef]
144.
Wang, S.; Wang, H. Teaching Higher Order Thinking in the Introductory MIS Course: A Model-Directed Approach. J. Educ. Bus.
2011,86, 208. [CrossRef]
145.
Groen, E.; Heijungs, R. Ignoring correlation in uncertainty and sensitivity analysis in life cycle assessment: What is the risk?
Environ. Impact Assess. Rev. 2017,62, 98–109. [CrossRef]
Sustainability 2024,16, 6364 22 of 22
146.
Hayek, F.A. Droit, Législation et Liberté: Une Nouvelle Formulation des Principes Libéraux de Justice et d”Économie Politique. Volume 3,
L’ordre Politique D’un Peuple Libre, 2nd ed.; Libre Échange, 0292-7020; Presses Universitaires de France: Paris, France, 1989.
147. Giddens, A. Les Conséquences de la Modernité; Torrossa: Fiesole, Italy, 1985; pp. 1–186.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.
