ArticlePDF Available

Uncertainties in macroeconomic assessments of low-carbon transition pathways -The case of the European iron and steel industry

June 2020
Ecological Economics 172(2020):106631

DOI:10.1016/j.ecolecon.2020.106631

License
CC BY 4.0

Authors:

Gabriel Bachner

University of Graz

Jakob Mayer

Karl Steininger

University of Graz

Annela Anger

University of Cambridge

Show all 6 authorsHide

The climate targets agreed in Paris 2015 render deep decarbonization of energy-and emission-intensive industries crucial. Policy makers are interested in the macroeconomic consequences of such decarbonization pathways and often rely on integrated modelling studies. However, the underlying modelling assumptions and uncertainties often remain unquestioned or invisible, although they may govern the models' results. For the case of a zero-process emission pathway of the European iron and steel industry, we demonstrate how different assumptions on different "layers" of uncertainty influence results. We show that effects strongly depend on technology choice, prevailing macroeconomic states as well as regional characteristics. The underlying socioeconomic development and the climate policy trajectory seem to play a less important role. Particularly, we find that the choice of model, i.e. which macroeconomic theory strand it arises from, influences the sign and magnitude of macroeconomic effects and thus should be well understood in terms of appropriate interpretations by both modelers and policy makers. We emphasize that model assumptions should be transparent, results sought as to be robust across a range of possible contexts, and presented together with the conditions under which they are valid. To that end, co-design and co-production in research would support its relevance and applicability.

Uncertainty ranges of EU-wide GDP and welfare change, relative to baseline, across scenario and model results (main scenario result levels are indicated by red dashed line). Left/right columns give high/low technology cost specification. (* = Excluding Greece, as no technology switch is implemented in this region.) (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

…

Figures - uploaded by Gabriel Bachner

Content may be subject to copyright.

Content uploaded by Gabriel Bachner

Content may be subject to copyright.

Contents lists available at ScienceDirect

Ecological Economics

journal homepage: www.elsevier.com/locate/ecolecon

Uncertainties in macroeconomic assessments of low-carbon transition

pathways - The case of the European iron and steel industry

G. Bachner

a,⁎

, J. Mayer

, K.W. Steininger

a,b

, A. Anger-Kraavi

, A. Smith

, T.S. Barker

c,e

Wegener Center for Climate and Global Change, University of Graz, Brandhofgasse 5, 8010 Graz, Austria

Department of Economics, University of Graz, Universitätsstraße 15/F4, 8010 Graz, Austria

Cambridge Econometrics, Covent Garden, Cambridge CB1 2HT, UK

Cambridge Institute for Sustainability Leadership, University of Cambridge, 1 Trumpington St, Cambridge CB2 1QA, UK

School of Environmental Sciences, University of East Anglia, Norwich NR4 7TJ, UK

ARTICLE INFO

Keywords:

Climate change mitigation

Uncertainty

Low carbon transition

Iron and steel

Macroeconomic modelling

Process emissions

ABSTRACT

The climate targets agreed in Paris 2015 render deep decarbonization of energy- and emission-intensive in-

dustries crucial. Policy makers are interested in the macroeconomic consequences of such decarbonization

pathways and often rely on integrated modelling studies. However, the underlying modelling assumptions and

uncertainties often remain unquestioned or invisible, although they may govern the models' results. For the case

of a zero-process emission pathway of the European iron and steel industry, we demonstrate how diﬀerent

assumptions on diﬀerent “layers”of uncertainty inﬂuence results. We show that eﬀects strongly depend on

technology choice, prevailing macroeconomic states as well as regional characteristics. The underlying socio-

economic development and the climate policy trajectory seem to play a less important role. Particularly, we ﬁnd

that the choice of model, i.e. which macroeconomic theory strand it arises from, inﬂuences the sign and mag-

nitude of macroeconomic eﬀects and thus should be well understood in terms of appropriate interpretations by

both modelers and policy makers. We emphasize that model assumptions should be transparent, results sought as

to be robust across a range of possible contexts, and presented together with the conditions under which they are

valid. To that end, co-design and co-production in research would support its relevance and applicability.

1. Introduction

By mid-century, the net balance of annual greenhouse gas (GHG)

emissions needs to be at least neutral to meet the climate targets as

deﬁned in the Paris Agreement. This demands fundamental changes in

the socio-economic systems (Rockström et al., 2017), which might

trigger a multitude of economy-wide eﬀects that are in synergy or

conﬂict with other societal objectives (e.g. full employment). Due to the

complexity of the socio-economic system, these eﬀects are surrounded

by a broad spectrum of uncertainties, which deserves attention (Aldred,

2012;Roelich and Giesekam, 2019).

In 2015, for UNFCCC Annex I countries, the highest share of GHG

emissions related to the combustion of fossil fuels (81%) (UNFCCC,

2017). A further, non-negligible, portion was comprised of industrial

process emissions (7%).

Thus, to achieve the goals as set out in the

Paris Agreement, industrial process emissions also need to be reduced

(if demands on negative emissions should not be excessive).

Considering that all sectors where process emissions occur are sub-

stantially trade exposed and that process emissions often account for

the dominant share of emissions of these sectors, the required process

emission reduction implies a signiﬁcant potential macroeconomic risk.

Nevertheless, compared to the analysis of combustion-based emissions,

only a narrow strand of literature focuses on the macroeconomic im-

plications of measures to tackle industrial process emissions (examples

are Bednar-Friedl et al., 2012;Pang et al., 2014;Schinko et al., 2014).

In our analysis, we contribute to this crucial strand of the literature and

focus on the economy-wide eﬀects of the low-carbon transition of the

European iron and steel sector. As the steel sector is deeply interwoven

in modern economies and expected to play a key role in the low-carbon

transition as supplier of high-tech materials, many factors might co-

determine how and in which direction the economy-wide eﬀects ma-

terialize. Thus, the iron and steel sector serves as an excellent case for

the analysis of risks and uncertainties as unintended outcomes might be

severe and manifold. In this article we build upon the analysis of Mayer

https://doi.org/10.1016/j.ecolecon.2020.106631

Received 11 April 2019; Received in revised form 5 February 2020; Accepted 20 February 2020

⁎

Corresponding author.

E-mail address: gabriel.bachner@uni-graz.at (G. Bachner).

The remaining shares refer to Agriculture (8%), Waste (3%) and other activities.

Ecological Economics 172 (2020) 106631

(http://creativecommons.org/licenses/BY/4.0/).

et al. (2019) who indicate the presence of large uncertainties, which we

here explore in more depth.

We introduce and operationalize diﬀerent layers of uncertainty (for

technology, socio-economics, policy, and macroeconomics), each of

which represents incomplete knowledge of diﬀerent aspects of the

problem at hand. Special emphasis is placed on the macroeconomic

layer, where we compare the results of two diﬀerent macroeconomic

modelling approaches. Speciﬁcally, we run a computable general

equilibrium (CGE) model (in spirit “neo-classical”) and a Post

Keynesian macro-econometric (PKME) model in their standard-setups.

Thereby we reveal methodological uncertainty, but also uncertainty at

the science-policy interface, since the commissioning of modelling

studies happens routinely and potentially without much reﬂection on

each modelling team's attribution to a speciﬁc school of thought. This in

turn might co-determine results and policy recommendations.

The core goals of this paper are: ﬁrst to pinpoint those layers which

drive uncertainties concerning the magnitude and direction of

economy-wide eﬀects of low-carbon transitions; and second to relate

these uncertainties to each other. To the authors' best knowledge, a

systematic and comprehensive analysis of uncertainties, as it is carried

out here, has not been published so far.

The remainder of the paper is structured as follows. In Section 2 we

comprehensively review the literature and clarify our contribution. In

Section 3 we develop the uncertainty framework, its layers and describe

the methods and assumptions to operationalize them. In Section 4 we

present results and discuss them. In Section 5 we draw conclusions.

2. Concepts of uncertainty in extant macroeconomic research

The literature oﬀers many deﬁnitions of uncertainty. One used

prominently was given by Knight (1921), who emphasizes that un-

certainty is an immeasurable lack of knowledge, whereas risk is mea-

surable via probability with known possible outcomes. In diﬀerent

schools of economic thinking, however, uncertainty is interpreted and

treated diﬀerently, e.g. in Keynesian economics there is an emphasis on

immeasurable uncertainty. In A Treatise on Probability Keynes distin-

guishes between the “weight”and the “probability”of an argument

(Keynes, 1921), with the former being described as “the degree of

completeness of the information”(Runde, 1990, p. 276). Sixteen years

later

he explicitly framed “uncertainty”in his General Theory as matters

for which “there is no scientiﬁc basis on which to form any calculable

probability whatever. We simply do not know.”(Keynes, 1937, p. 214). For

Keynes uncertainty is an essential cause for “hoarding money”(thus for

liquidity), leading to reluctant investment, spare production capacities

and involuntary unemployment (Lucarelli, 2010). Inﬂuential Post

Keynesian economists such as Shackle (1943; cit. in Zappia, 2014), or

later Davidson (1991) followed this thinking and thus Keynesian un-

certainty is still very salient for Post Keynesian economists. A more

recent interpretation of this causality is that “fundamental uncertainty”

makes it impossible for agents to use all available resources fully

(Mercure et al., 2019), leading to idle available resources. Another early

statement on uncertainty in (Ecological) economics was given by

Georgescu-Roegen (1954, p. 524), referring to Knight and Keynes: “In

the case of risk, but not in the case of uncertainty, we can deﬁne the

probability of the outcome.”Neoclassical economics, on the other hand,

treats probabilistic risk and uncertainty synonymously; expectations are

at the center, which are based on past data and respective probabilities

or on subjective perceptions on probabilities. From a mainstream per-

spective the demand for liquidity is thus irrational (Davidson, 1991;

Machina, 1987) and money is neutral.

The importance of the diﬀerent treatment of uncertainty in diﬀerent

economic schools is emphasized by Mercure et al. (2019, p. 1023):“The

theoretical diﬀerence between the schools has at its heart a diﬀerence in the

treatment of risk and uncertainty”and Davidson (1991, p. 141) concludes

that “The analyst must (…)choose which system is more relevant for

analyzing the economic problem under study.”Hence, uncertainty plays

diﬀerent roles in diﬀerent economic schools and co-determines how

respective models work, but there is no agreed handling of uncertainty

in economics.

So far we have touched upon how uncertainty is treated within

economics,

however, uncertainty also plays an important role at the

science-policy interface. This is emphasized by the IPCC (Kunreuther

et al., 2014, p. 155), who deﬁnes uncertainty as “a cognitive state of

incomplete knowledge that results from a lack of information and/or from

disagreement about what is known or even knowable.”They further dif-

ferentiate three diﬀerent types of uncertainty. First, epistemic un-

certainty emerges from the “lack of information or knowledge for char-

acterizing phenomena.”Second, paradigmatic uncertainty results “from

the absence of prior agreement on the framing of problems, on methods for

scientiﬁcally investigating them, and on how to combine knowledge from

disparate research traditions.”Third, translational uncertainty “results

from scientiﬁc results that are incomplete or conﬂicting so that they can be

invoked to support divergent policy positions.”(ibid., p. 178).

For the operationalization of uncertainty in empirical economic

research, the IPCCs typology might be too blurry, though. For example,

the choice between a Neoclassical or Post Keynesian macroeconomic

model would be subject to uncertainty on two issues: how to char-

acterize phenomena (epistemic uncertainty) and the absence of agree-

ment on how to investigate a problem (paradigmatic uncertainty).

Moreover, in the literature epistemic uncertainty is typically contrasted

with (irreducible) aleatoric uncertainty or variability (see e.g. Basu,

2017;Bjarnadottir et al., 2019;Morales-Torres et al., 2019;Walker

et al., 2003), with epistemic uncertainty resulting from not enough

knowledge to properly describe a system (parameters but also func-

tional relationships), whereas aleatoric uncertainty results from ran-

domness. Using this typology, macroeconomic model uncertainty is

clearly epistemic for those models which are currently predominantly

used (deterministic models), but might be also aleatoric (or chaotic)

when other approaches (such as system dynamics) are chosen.

For managing uncertainty in model-based decision support Walker

et al. (2003) suggest a harmonized framework. They deﬁne uncertainty

as “any deviation from the unachievable ideal of completely deterministic

knowledge of the relevant system”(ibid., p. 5). The authors emphasize

that uncertainty should be regarded along three dimensions: location

(system boundary, model parameters and variables, functional re-

lationships), level (spectrum known to unknown) and nature (epistemic

versus variability). Refsgaard et al. (2007) further extend this frame-

work and suggest suitable methodologies for addressing diﬀerent types

of uncertainty within the various dimensions.

Another –in our view very helpful –typology for uncertainty for

applied policy-related science is given by Reilly and Willenbockel

(2010), who refer to the seminal work by Funtowicz and Ravetz (1990),

and diﬀerentiate between technical, methodological and epistemolo-

gical uncertainty. Technical uncertainty refers to data quality for model

calibration and the assumption for the drivers of change, methodolo-

gical uncertainty refers to the structure, functional forms and beha-

vioral equations of the model, and epistemological uncertainty refers to

changes in human behavior and values, technological surprises and so

called high impact-high uncertainty events, but also to deeper un-

certainty such as randomness and surprises. Funtowicz and Ravetz

(1990, p. 38) state that “the more complex aspects of methodological

Although it is questionable whether Keynes followed his early theory of

probability from 1921 when writing his General Theory, which was published in

1936 (O'Connell, 1996).

Since climate policy assessment is currently strongly based on Neoclassical

computable general equilibrium models, with Post Keynesian econometric

input-output models gaining traction, the focus in this paper is on these two

schools of thought.

G. Bachner, et al. Ecological Economics 172 (2020) 106631

uncertainty, and epistemological uncertainty as well, are outside the

boundaries of any calculi.”, thus going even beyond aleatoric uncertainty

or variability, implicitly referring to immeasurable uncertainty in the

Knightian sense.

For managing uncertainty, there are several approaches.

Measurable uncertainty is usually addressed via probabilistic meth-

odologies such as Monte Carlo simulations. More challenging is the

management of immeasurable uncertainty. For that, a powerful ap-

proach is the creation of “alternative futures”(Dror, 1970) or scenario

analysis (Refsgaard et al., 2007;Reilly and Willenbockel, 2010). Also

Gowdy and Erickson (2005) suggest that scenario analysis with tech-

nical descriptions of particular economies should be used to address

uncertainty (a “structural approach”;Duchin, 1998 cit. in Gowdy and

Erickson, 2005), and propose the application of the precautionary

principle. Methodological uncertainty and uncertainty about functional

relationships or behavior (epistemic/epistemological) can be tackled

via multi-model comparisons.

We here aim to push the boundary of uncertainty appraisal in em-

pirical macroeconomic policy-related research by going into deeper

dimensions of uncertainty as it is usually done and to demonstrate that

there are also major uncertainties at the science-policy interface. Our

approach is motivated by the existing empirical literature on low-

carbon transitions, where uncertainty appraisal is mostly done via

scenario analysis, e.g. Fujimori et al. (2014) combine scenario as-

sumptions on mitigation eﬀorts and energy demand. An example for

Monte Carlo simulations is given by Abler et al. (1999), who investigate

economic and sectoral policies impacting environmental indicators, or

by Babonneau et al. (2012) and Wang and Chen (2006) focusing on

climate policies. An example of multi-model comparison is given by the

Energy Modelling Forum 29 (Böhringer et al., 2012), however all of the

models used are general equilibrium models, which all share the same

basic functional mechanism.

The same is true for an analysis by

Guivarch et al. (2016), who test for uncertainty across diﬀerent socio-

economic developments.

To summarize, and using the typology of Funtowicz and Ravetz

(1990), current approaches most often do not go beyond technical

uncertainty. Yet, for a fully-ﬂedged assessment, it is essential to include

methodological, and ideally epistemological, uncertainty. In the deﬁ-

nition framework of Walker et al. (2003) this means to not only capture

the location dimension of uncertainty, but also including the nature

dimension. This ultimately leads to discussions on fundamental (in our

case macroeconomic) questions, such as whether the economy is

supply- or demand-driven, and which model to choose in which situa-

tion. This discussion is not new. For example already Dutt (1984)

pointed out that the policy implications on income distribution are very

diﬀerent when assuming either a “supply-driven”or a “demand-driven”

model. Yet, this dimension of uncertainty has so far not been carried

over to the ﬁeld of climate change economics as modelling groups of

diﬀerent schools, such as Neoclassical and Post Keynesian, mostly work

in parallel strands, rather than together. Notable exceptions, and thus

attempts for such an integrated analysis are oﬀered by Edenhofer et al.

(2010),Jansen and Klaassen (2000),Kober et al. (2016) or Meyer and

Ahlert (2019), who all implicitly address this type of uncertainty (as

each of them deploys diﬀerent macroeconomic model types). However,

the mentioned studies do not explicitly embed their ﬁndings in a

broader framework of uncertainties –and most importantly, do not

oﬀer a comparison across diﬀerent types. Only recently Mercure et al.

(2019) conceptually discuss the diﬀerences between general equili-

brium models and non-equilibrium models and the implications for

(climate) policy; however, they only mention preliminary results of a

modelling exercise in which two model of diﬀerent type have been

enhanced regarding innovation and ﬁnance, with convergence of

models as a result. Our research contributes to this fundamental dis-

cussion on ecological macroeconomics (Fontana and Sawyer, 2016;

Rezai and Stagl, 2016).

3. Uncertainty layers, methods and scenarios

In this section we lay out our methodological framework. Our

general approach is to combine diﬀerent “uncertainty layers”to create

diﬀerent worlds in which a switch from a baseline technology (i.e.

process emission-intensive steel production) to a process emission-free

alternative takes place (see Fig. 1). The technology switch itself gen-

erates (i) the technological uncertainty layer, where a low and a high-cost

technology speciﬁcation span the uncertainty range on costs. The

second layer (ii) captures socio-economic uncertainty, with variations

across shared socio-economic pathways (SSPs). Third (iii), the climate

policy uncertainty varies the stringency and geographic scope of climate

policy. Fourth (iv), for the macroeconomic uncertainty layer we apply

diﬀerent macroeconomic model types, thereby changing the assump-

tion of the prevailing macroeconomic state and also capturing un-

certainty at the science-policy interface. The speciﬁc assumptions of the

four layers are explained in Sections 3.1-to 3.4. To connect our meth-

odology to the various typologies of uncertainty as discussed in Section

2,Table 1 assigns the layers to diﬀerent types of uncertainty and brieﬂy

explains how these uncertainties are implemented in our modelling

framework. Note, that in Table 1 we also include “other layers”,aswe

do not want to claim completeness and because we treat translational

uncertainty qualitatively when discussing our results. Relating our ap-

proach to the framework of Refsgaard et al. (2007), the quantitative

part of our analysis combines multiple model simulations (“having dif-

ferent conceptual models based on diﬀerent (…)interpretations”; ibid., p.

1550) and scenario analysis (“deal explicitly with diﬀerent assumptions

about the future”; ibid., p. 1551).

3.1. Technological uncertainty layer

There exist several technologies to produce process emission-free

iron and steel, opening up the ﬁrst uncertainty layer. We specify two

alternative technologies, replacing the current CO

-emission-intensive

blast-furnace basic oxygen furnace route (BF-BOF). First, hydrogen-

based direct reduced iron, which is fed into an electric arc furnace (DRI-

H-EAF). This technology is mature and already in use, yet not based on

hydrogen (H), but natural gas (CH

). Second, plasma-direct-steel-pro-

duction (PDSP); a single-step production process, with iron ore as the

only relevant raw material input. PDSP is still in the experimental

phase, but seems promising with respect to unit costs, ﬂexibility of

scale, steel quality and GHG emissions (Sabat and Murphy, 2017).

By analyzing present estimates of production cost range of a mature

technology and one which is still in its infancy, we cover a broad range

of uncertainty. The respective operation cost structures are given in

Table A1. To broaden the cost range we also assume diﬀerent industry

speciﬁc electricity prices, leading to two techno-economic speciﬁca-

tions: High-cost (i.e. DRI-H-EAF with electricity costs of 0.05 EUR/kWh)

or low-cost (i.e. PDSP with 0.03 EUR/kWh). Thus, these two speciﬁca-

tions presumably represent the extreme ends of the cost spectrum (at

least according to the current knowledge) capturing as much of the

uncertainty regarding unit costs as possible. Note, that the low-cost

speciﬁcation has still higher unit costs than the conventional tech-

nology (BF-BOF), thus representing the low-cost alternative speciﬁca-

tion. Also note, that during the investment phase, additional capital

expenditures (annuities for new facilities) increase the unit cost of the

new technologies (on top of the already higher operating costs). For the

upscaling of the PDSP technology, further research and development (R

&D) is required; how much is uncertain, though. We thus follow a

At least if diﬀerent theories and corresponding models for these relation-

ships or behaviors are available.

There are some models which allow for out of equilibrium conditions on

certain markets (e.g. the labor market), but the main behavioral assumptions of

optimization under budget/resource constraints are still the same.

G. Bachner, et al. Ecological Economics 172 (2020) 106631

pragmatic approach and assume that these R&D investments are ﬁ-

nanced publicly, but crowding out other R&D investment elsewhere, i.e.

without an impact on the absolute level of economy-wide investment.

For technological and cost details see the Appendix.

In our analysis the deployment of the new technologies is de-

terministic, i.e. exogenous, which we interpret as a zero-emission

standard. The industry thus cannot simply decide to pay the carbon

price and keep the conventional technology.

3.2. Socio-economic uncertainty layer

To capture socio-economic uncertainties, we deploy diﬀerent SSPs,

which are the standard scenarios in the climate change research com-

munity. SSP2 (O'Neill et al., 2017, 2014)isa“Middle of the road”socio-

economic background development, with moderate climate change

mitigation and adaptation challenges. We deploy additionally SSP3 and

SSP5, which span the uncertainty range of regional economic growth

rates for EU regions (see Fig. A2). The narrative of SSP3 is characterized

by regional rivalry, which unfolds in slow economic growth and slow

technological change, whereas SSP5 is characterized by strong growth

and excessive (fossil) energy use. Compared to SSP2, the SSPs 3 and 5

imply higher challenges for mitigation (which we model by lower

multi-factor productivity; see Fig. A2). SSP3 also implies higher chal-

lenges for adaptation due to low adaptation capacities (low income),

whereas SSP5 is characterized by low adaptation challenges. To capture

these diﬀerences, we mimic high/low adaptation challenges by in-

creasing/decreasing capital depreciation (by ± 25%, relative to SSP2).

This should serve as a representation of higher or lower vulnerability to

climate change and thus diﬀerences in capital stock accumulation.

The

long-term capital depreciation rates we use are given in Table A2.

3.3. Climate policy uncertainty layer

For the climate policy layer, we specify three diﬀerent climate-

policy settings. First, reluctant policy reﬂects a modest global CO

price,

reaching 46 EUR

2011

/tCO

globally by 2050 (based on IEA (2016); see

Fig. 2). Second, ambitious policy reﬂects a more stringent policy by tri-

pling the reluctant CO

price trajectory, so it reaches 138 EUR

2011

/tCO

globally by 2050. This is in line with the Interagency Working Group on

Social Cost of Carbon (IWGSCC, 2015)

and the 450 ppm scenario by

the International Energy Agency (IEA, 2016).

The ambitious policy

world thus may reﬂect an upper bound of politically feasible CO

pri-

cing. Third, we set up an EU-ambitious case in which the EU (within its

emission trading scheme, ETS) implements an ambitious policy but the

rest of the world remaining reluctant.

3.4. Macroeconomic uncertainty layer

3.4.1. Macroeconomic models: full capacity utilization vs. available

resources

At the macroeconomic uncertainty layer we use two macro-

economic model types, being based upon diﬀerent underlying macro-

economic theories. One model is a computable general equilibrium

(CGE) model, which describes the economy in a state of full capacity

utilization. The basic idea behind CGE models is that all markets are in

equilibrium (i.e. supply = demand) and this “general equilibrium”can

be disturbed by an intervention (e.g. by an enforced switch to a new

technology), triggering relative price changes as well as demand

(quantity) adjustments until a new general equilibrium emerges. From

the diﬀerence between new and old equilibrium we draw conclusions

on how the economy reacts to the intervention. In recursive dynamic

CGE models (which are predominantly used), such annual equilibria are

extrapolated into the future via the connection of investment and ca-

pital accumulation over time. Hence, the investment activity in one

year, determines the capital availability of the next year. The adjust-

ment process that leads to the new equilibrium happens smoothly and

instantaneously and is driven by assumptions on the behavior of eco-

nomic agents. Producers are assumed to maximize proﬁts and con-

sumers to maximize utility out of consumption, subject to prices, factor

and income availability as well as ﬂexibility in substitutability across

factors and goods.

As optimal behavior is assumed, all available pro-

duction factors are fully used (factor prices adjust until this is the case).

This process reﬂects the long-term optimality perspective of CGE

Baseline

technology

Process emission-

free technologies

Macroeconomic state

Climate policy

Socio-economic development (SSP)

Low-

cost

High-

cost

“World” in which technology switch takes place

Fig. 1. Uncertainty framework and diﬀerent layers of uncertainty, which in combination create the world in which the technology switch to a process emission-free

iron and steel technology takes place.

We assume that the physical climate change scenario is the same across

socio-economic developments, however we do not implement a speciﬁc tem-

perature trajectory with respective economic damages.

Social Costs of Carbon of USD

2011

212/tCO

in 2050 (95th percentile)

Reaching USD

2011

140/tCO

in 2040.

In quantitative modelling this ﬂexibility is typically speciﬁed via elasticities

of substitution; a feature that sets CGE models apart from Input-Output or ba-

sically also econometric models, which assume ﬁxed material production

structures.

G. Bachner, et al. Ecological Economics 172 (2020) 106631

models, implying that cost changes fully pass through to ﬁnal demand.

The main characteristics of CGE models are thus that relative price

changes drive the system towards a new equilibrium and that factor

supply constrains economic activity; i.e. there are no idle capacities, the

economy runs at (optimal) full capacity utilization and growth is driven

by factor supply changes (the model is “supply-driven”).

In addition, we deploy a Post Keynesian macro-econometric (PKME)

model that allows for economies not working at full capacity utilization

and for rigidities in adjustment processes. In such models, economic

growth is “demand-led”, meaning that demand changes drive the

growth of output in the model. There is no concept of the economy

being in a general, or full-employment, equilibrium. Industries are not

assumed to behave optimally, but the model simulates behavior as

derived from past observation (using econometric estimation). Thus,

typically technological coeﬃcients are ﬁxed and no input substitution

in production takes place. In case of additional demand stimulus,

available resources can be activated, which would increase economic

output and income. In contrast to CGE models, PKME models are mo-

tivated by both short-run and long-run perspectives and thus explicitly

capture the transition between the past and the future (non-equilibrium

states), whereas CGE models compare depictions of the economy that

are in equilibrium, before and after a system intervention takes place

(long-run perspective). This also means that in PKME there may be full-

cost pass-through to ﬁnal demand only in the long run.

To summarize, general diﬀerences between CGE and PKME are re-

ﬂected by diﬀerent assumptions on the degree of capacity utilization

(full in CGE versus idle/unknown in PKME) as well as the way in which

economic behavior is treated (optimization in CGE versus based on

historical behavior in PKME). Further, CGE models are supply-driven,

assuming full employment, whereas PKME models are demand-driven,

constrained by full employment and policies preventing excessive in-

ﬂation. These two model classes show the extreme ends of the spec-

trum. An example for a “hybrid”New Keynesian model, in which op-

timization and econometric input-output modelling is combined is

given by Sommer and Kratena (2017).

3.4.2. The WEGDYN model

For the case of average long-term full capacity utilization we deploy

the Wegener Center Dynamic CGE model WEGDYN (Mayer et al.,

2018), which is a global multi-sector, multi-country, recursive dynamic

CGE model. WEGDYN is based on the GTAP9 database (Aguiar et al.,

2016). The model solves for static equilibria in ﬁve-year steps,

Table 1

Overview of the uncertainty layers used in the analysis, respective type of uncertainty that is addressed and model implementation.

Uncertainty layer Typology and respective type of uncertainty captured

Model implementation via

Knight (1921),Keynes (1921, 1937) IPCC, Kunreuther et al.

(2014)

Funtowicz and Ravetz (1990)

Technology No: some costs estimates of technologies are

available

Epistemic Technical Two diﬀerent iron and steel production technologies

Socio-economic No: plausible corridors for socio-economic

development until 2050 are given

Epistemic Technical Three diﬀerent shared socio-economic pathways (SSPs)

Climate policy Yes: How climate policy might evolve is uncertain

Epistemic Epistemological Three diﬀerent climate policy worlds (stringency and geographic scope)

Macroeconomic state Yes: We do not know, how macroeconomic states

will look like in the future

Epistemic/paradigmatic Methodological/

epistemological

Two diﬀerent macroeconomic models, with main diﬀerences being in assumptions about degree of

capacity utilization and the speed of adjustment

Other layers Translational Not directly captured in the here applied models, but using two diﬀerent macroeconomic models

opens up the discussion on the quantiﬁcation of translational uncertainty at the science-policy

interface (cf. Section 4.4).

Yes/no for Knight/Keynes describes whether our layers capture Knightian/Keynesian uncertainty or not.

See e.g. sudden change due to presidency of Donald Trump in the USA.

100

120

140

160

2015 2020 2025 2030 2035 2040 2045 2050

€2OCt/

reluctant ambitious

Fig. 2. Diﬀerent CO

price trajectories [EUR

2011

/tCO

These relative prices include wage, interest and exchange rates, but there

are also CGE models which ﬁx certain prices, or assume their one-direction

rigidity, such as downward rigidity of the real wage rate to model minimum

wages and unemployment.

G. Bachner, et al. Ecological Economics 172 (2020) 106631

connected through endogenous capital stock accumulation, exogenous

labor force (population) growth and multi-factor productivity growth.

Activities of production and consumption involve the emission of

carbon dioxide (CO

) originating from the combustion of fossil fuels

and/or from industrial processes. For this analysis, WEGDYN is set up

with 15 economic sectors (Table A3). For the iron and steel sector we

diﬀerentiate between three sub-sectors representing: (i) a conventional

and process emission-intensive technology (BF-BOF), (ii) a new process

emission-free technology (as discussed in Section 3.1) and (iii) the

casting, rolling and ﬁnishing of crude steel. The former two sub-sectors

supply crude steel, which is combined with output of the third sub-

sector to provide a ﬁnal iron and steel market supply. WEGDYN divides

the world into 16 regions (Table A4), with a focus on the EU.

On the

labor market, classical unemployment is implemented via a minimum

wage; hence this market is not in equilibrium. Foreign trade is im-

plemented via bilateral trade ﬂows on a sectoral basis, where domes-

tically produced goods are imperfect substitutes for foreign goods

(Armington, 1969). Changes in trade patterns are triggered by relative

prices changes. Price formation is thus a function of unit cost changes,

endogenous relative price eﬀects on the domestic markets as well as

import price eﬀects. Please see Appendix A2 for additional information.

For the full model documentation see Mayer et al. (2018).

3.4.3. The E3ME model

For the case of a Post Keynesian macroeconomic treatment of the

economy, we use the global Energy-Environment-Economy Macro

Econometric model E3ME (for application see Barker et al., 2012).

E3ME is a PKME model in which functional relationships are de-

termined by parameter estimations for 28 sets of short and long-term

econometric equations at regional and sectoral level, using historical

data since 1970. However, the input-output coeﬃcients are ﬁxed for

each year (on exogenous technological trends), except those for the

energy inputs, which are derived from an energy module allowing for

substitution when relative prices change. As opposed to CGE models,

E3ME includes economies of scale. E3ME allows for frictions and time

lags, with investment and employment being determined by each in-

dustry in each country, depending on expected outputs, and the prices

and wage costs per unit of output in relation to the output prices. Most

components of total demands and the divisions between outputs and

imports are aﬀected by relative prices at disaggregated level.

Proxies for the utilization of capacity for each industry in the pro-

jections are given by the ratio of output (solved from demand) to ex-

pected or “normal”output (from autoregressive equations on past

output), although this converges to unity in the long run. In E3ME

money supply is endogenous, hence there are spare capacities on the

capital market and thus there is no full crowding out of investments

(Pollitt and Mercure, 2018). The second key spare capacity is in the

labor market. The available labor force depends on participation rate

equations using exogenous population forecasts. The proxies for capa-

city utilization are especially important for the short-run responses: a

higher ratio leads to more inﬂationary pressure, all else being equal.

The main supply-side constraint is the availability of labor (labor par-

ticipation rate is endogenous). Production levels exceeding expected

levels lead to higher prices and higher rates of import substitution.

There is an extensive dynamic treatment of prices, wages and costs

and the eﬀect of changes in relative prices on the real economy in

E3ME. The full equation for price formation in E3ME includes unit

costs, import prices (capturing international competition eﬀects),

technology indices, and the ratio of actual to expected output. The

degree to which exogenous cost changes are passed through to prices is

driven by the level of competition in the sector, determined by the

parameter for unit costs in the econometric equation. In the long run,

the assumption of ‘Invariance of Tax Incidence’is imposed on the es-

timated equations, so that all price and unit-cost changes in total sup-

plies (i.e. including imports) are passed on to ﬁnal demand. Barker and

Gardiner (1996) provide a full description of the estimated employ-

ment, wage and price equations in an earlier EU-only version of E3ME.

Please see Appendix A3 for additional information. For the full model

documentation see Cambridge Econometrics (2019).

3.4.4. Similarities between WEGDYN and E3ME

Despite the fundamental diﬀerences, WEGDYN and E3ME also have

similarities, namely, the high level of sector disaggregation and sector

interconnectedness via an input-output framework, that makes both

approaches capable of examining varying sectoral impacts, both di-

rectly and indirectly (through changes in demand from other sectors

and changes in disposable incomes). These economy-wide –as opposed

to partial –approaches recognize implications for sectors that are not

directly targeted by a particular intervention in the economic system,

such as a GHG emission mitigation measure. Also, both models capture

foreign trade via bilateral trade and international competition, which

inﬂuences price formation on domestic markets. Model speciﬁcdiﬀer-

ences are presented with the discussion of our results.

3.5. Scenario deﬁnitions and comparisons

To address the ﬁrst three uncertainty layers (technology, socio-

economic, climate policy) we make use of scenario analysis. Fig. 3 ex-

tends Fig. 2 and shows how the variations on the diﬀerent layers are

combined into scenarios. In each of the dashed boxes, the technology

switch to either a high-cost or a low-cost alternative takes place and is

compared to the baseline. This diﬀerence is indicated by a delta sign

(Δ), which is however diﬀerent in each box, as the switch happens

under diﬀerent circumstances, capturing the uncertainties. In the re-

sults, we thus compare diﬀerences in diﬀerences. As a starting point, we

construct a main scenario (MAIN), which is described as follows. For

the socio-economic layer we assume SSP2, for the climate policy layer

we assume a globally reluctant climate policy and for the macro-

economic layer we assume full capacity utilization (by using WEGDYN).

We then deviate from MAIN according to the arrows in Fig. 3. The

quantitative analysis aims at tackling technical and epistemic/episte-

mological uncertainties regarding technologies, policies and the socio-

economic future. The deployment of fundamentally diﬀerent macro-

economic models, WEGDYN (full capacity utilization) and E3ME (with

idle available resources), allows us to reveal epistemic/paradigmatic

(Kunreuther et al., 2014) or epistemological/methodological un-

certainty (Funtowicz and Ravetz, 1990). At this point we need to stress,

that our quantitative analysis does not directly capture translational

uncertainty, however, as we use diﬀerent macroeconomic models in

their standard setups we make a ﬁrst step towards reducing transla-

tional uncertainty at the science-policy interface.

4. Results and discussion

4.1. Technological and socio-economic uncertainty

We ﬁrst report results for the iron and steel sector itself, being at the

very core of the system intervention. All results are given relative to a

respective baseline, in which the new iron and steel technologies are

not activated and all other aspects remain equal. Note, that we initially

assume full capacity utilization (i.e. we use the WEGDYN model).

Fig. 4 shows eﬀects on regional market prices for iron and steel. The

top/bottom row gives results for the high/low-cost technology speciﬁ-

cation. The diﬀerence between top and bottom row thus represents the

technological uncertainty. In addition, we depict the socio-economic

Actually, EU + 3 (EU28 plus Norway, Lichtenstein and Iceland). Austria

(AUT) and Greece (GRC) are modeled as individual model regions, in order to

contrast implications of high process emission-intensive iron and steel pro-

duction of the former country with no process emissions-intensive production of

the latter (see Fig. A1).

G. Bachner, et al. Ecological Economics 172 (2020) 106631

Socio-economic development (SSP)

SSP2 SSP3 SSP5

Macroeconomic state

MAIN:

Full

capacity

Available

resources

Macroeconomic uncertainty

Full capacity

Full

capacity

Climate policy uncertainty

Full

capacity

Socio-economic uncertainty

ycilopetamilC

Globally reluctant

EU ambious

Globally ambious

Baseline

technology

Process emission-

free technology

Low-

cost

High-

cost

Techn. uncertainty

Fig. 3. Uncertainty space, emerging from the combinations of diﬀerent uncertainty layers (Technological, Socio-Economic, Climate Policy, Macroeconomic state).

2011

2021

2031

2041

2050

-6%

-4%

-2%

+2%

+4%

+6%

+8%

+10%

+12%

+14%

+16%

stekra

ni e

AUT EEU NEU SEU WEU GRC

high cost

2011

2021

2031

2041

2050

2011

2021

2031

2041

2050

2011

2021

2031

2041

2050

Change in m arkets price

2011

2021

2031

2041

2050

-6%

-4%

-2%

+2%

+4%

+6%

+8%

+10%

+12%

+14%

+16%

2011

2021

2031

2041

2050

rp ste

kra

egn

low cos t

SSP2 (main scenario)

SSP3

SSP5

Fig. 4. Change of iron and steel market prices, relative to the baseline with SSP variations on the socio-economic uncertainty layer (assuming reluctant climate policy

and full capacity utilization [WEGDYN]).

G. Bachner, et al. Ecological Economics 172 (2020) 106631

uncertainty layer by simulating the technology switch within two al-

ternative SSPs (SSP3 and SSP5, in addition to SSP2).

With a high-cost speciﬁcation, prices are above baseline levels in all

regions, with peaks ranging between +15% (Austria in 2031) and

+2% (Greece

in 2031) and declining thereafter, but remaining above

the baseline until 2050. Prices are higher due to higher production

(unit) costs, both because of additional capital costs due to investment

requirements (annuities for ﬁnancing new facilities), and because of

higher operating costs (Table A1).

With a low-cost speciﬁcation, the

price more than compensates the production (unit) cost dis-

advantage of the process emission-free technology in the long-term,

which leads to lower iron and steel prices in all regions in 2050, ranging

between −5% (Austria in 2050) and −1% (Greece in 2050). In the

short-term, however, also the low-cost speciﬁcation leads temporarily

to higher prices, especially due to capital investment costs.

Regarding regional diﬀerences, Austria seems to be the most sen-

sitive region. This is because the share of the BF-BOF steelmaking in the

iron and steel industry is highest in Austria (> 90% of steel is produced

along this route), whereas the shares in EEU (Eastern EU), NEU

(Northern EU) and WEU (Western EU) lie at 65% and in SEU (Southern

EU) only at 30% (see Fig. A1 for details).

Concerning socio-economic uncertainty, results are robust as the

direction of eﬀects does not vary across SSPs. However, in some regions

we see relatively large diﬀerences when comparing SSP3 to SSP2 or

SSP5. This is because in SSP3 depreciation rates are higher (see Section

3.2). Thus, capital prices are higher and as the new technology is more

capital intensive, price eﬀects are pushed upwards (as compared to

SSP2 and SSP5). In other words, when introducing a capital-intensive

technology in a world with high capital prices, its eﬀects are stronger

(to the positive in terms of price eﬀects) than in a world with lower

capital prices. The opposite eﬀect is the case for SSP5, where depre-

ciation rates are below the SSP2 rate.

The changes in relative prices translate into output eﬀects, which

are shown in Fig. 5a, aggregated to EU level. In 2050, with a high-cost

technology speciﬁcation, output is lower relative to the baseline (be-

tween −6% and −13%), whereas with a low-cost technology speciﬁ-

cation, output is higher (between +4% and +8%). Again, concerning

socio-economic uncertainty our results are robust in terms of direction

and magnitude, with stronger eﬀects under SSP3. For detailed regional

output eﬀects see Fig. A15.

The eﬀects from the iron and steel sector also have eﬀects on the

rest of the economy and eventually on GDP and welfare. EU-wide GDP

eﬀects are shown in Fig. 5b, again for the main scenario with variations

of SSPs. For the low-cost case we observe a lower GDP during the

transition by up to −0.8%, with potential long run GDP gains (+0.25%

in 2050). This can be explained as follows. In the ﬁrst periods of the

transition the higher unit-costs translate into lower productivity of the

iron and steel sector, which lowers GDP (compared to the baseline).

However, the unit cost disadvantage (w.r.t. baseline) disappears and

ﬂips into an advantage, due to the rising CO

price, hence productivity

increases. Thus, the GDP eﬀect turns positive, but not immediately, as it

needs some time to return to baseline levels via faster capital accu-

mulation (which is getting faster as soon as unit costs are below the

baseline level). With a high-cost technology speciﬁcation unit costs also

increase, leading to a loss of productivity and thus a lower GDP (−1.6%

by 2050). Even after the investment phase (beyond 2031), production

costs (and prices) are still higher and thus GDP does not return to the

baseline by 2050, but remains at a lower level (−1.5%). See Fig. A16

for region-speciﬁceﬀects.

Regarding socio-economic uncertainty, the results are robust in

terms of direction and magnitude for the high-cost speciﬁcation. For the

low-cost technology speciﬁcations, however, the GDP eﬀects vary in

their long-term (2050) eﬀects across SSPs, with slight negative eﬀects

under SSP3 and slight positive eﬀects under SSP2 and SSP5. This is

because under SSP3 capital prices are higher, which leaves the newly

introduced –more capital-intensive –technology with a stronger cost-

disadvantage compared to the baseline technology.

What is not captured in GDP is the shift from consumption to sav-

ings/investments for new production facilities. Since we assume full

capacity utilization in the main scenario, consumption is reduced to

oﬀset the additional investment. This reduction of consumption is re-

ﬂected in a lower welfare

level. The eﬀect is similar, however, more

severe than the GDP eﬀect, since GDP also includes the positive eﬀect of

higher investments. We observe lower welfare levels of up to −3% in

the high-cost case (in EEU in 2050) and higher welfare levels of up to

+0.5% in the low-cost case (in WEU in 2050; see Fig. A17 for details).

For employment, the eﬀect depends on the sign of the change in

labor intensity in combination with the overall productivity (unit cost)

change of the new technology. In our analysis, the new steelmaking

technologies are characterized by lower labor intensities, which in

isolation leads to decreased employment due to lower labor demand. In

addition, negative productivity eﬀects (in terms of higher unit costs)

adds to this eﬀect, paired and ampliﬁed with indirect eﬀects to other

sectors due to a strong dependency on steel (see Figs. A18 and A19 for

regional eﬀects). By 2050, unemployment rates are thus higher by up to

+2.5%-points (as compared to the baseline), when assuming the high-

cost technology speciﬁcation. For the low-cost technology speciﬁcation,

unemployment rates also tend to be above baseline levels, however the

eﬀects are less severe and unemployment tends to return to baseline

levels –or even below –in the long-run as overall productivity and

growth catch up.

4.2. Climate policy uncertainty

We now investigate the climate policy layer. Again, we compare to

the main scenario, but now for diﬀerent policy worlds, implemented as

diﬀerent exogenous CO

prices (in both the baseline and in the tech-

nology-switch scenarios). Fig. 5c illustrates EU-wide GDP eﬀects. With

a high-cost technology speciﬁcation, the highest GDP losses emerge in

the policy world with lowest support for clean technologies, i.e. the

“globally reluctant”world. In the “EU ambitious”world, the transition

is less costly, since there is more policy support for the process emis-

sion-free technology with a smaller production cost disadvantage. We

see that GDP losses are weaker (by up to 0.5%-points in 2050), when

the transition takes place in a world with a more ambitious climate

policy in the EU. EU-wide GDP losses are rather insensitive to non-EU

policy (see Fig. A20 for regional details).

With a low-cost technology

speciﬁcation, climate policy helps to foster the positive GDP eﬀect.

Compared to the “globally reluctant”world, ambitious EU policy can

increase GDP gains by +1%-point (in 2050) with slightly higher ben-

eﬁts in the “globally ambitious”world. In general, we see that the

Note that in Greece the eﬀect emerges only indirectly via international

trade, since in Greece itself there are no BF-BOF steelmaking facilities to be

replaced.

Note that in the baseline there would be 100% conventional steel output

over the full time horizon, however with a still rising CO

price, which would

increase production costs of the (baseline) BF-BOF technology, whereas pro-

duction costs for the process emission-free technology would start to fall in

2036. Hence, the relative cost-disadvantage of the process emission-free tech-

nology decreases from 2036 onwards and thus the prices of the baseline tech-

nology and the process emission-free alternative would converge, which leads

to an inverted U-shaped eﬀect.

Measured by means of Hicksian Equivalent Variation, which describes the

change in consumption possibilities, or the willingness to pay (accept) for a

price rise (fall) not to occur.

This can be explained via the model closure for foreign trade: In the

standard setup, we assume a ﬂexible exchange rate and ﬁxed foreign savings

(current account balance). Thus, regional trade patterns can change but not the

absolute level of net exports.

G. Bachner, et al. Ecological Economics 172 (2020) 106631

uncertainty from the climate policy layer inﬂuences the magnitude of

the GDP eﬀect; however, the sign of the eﬀect is robust. The same

patterns as observed for GDP apply to welfare eﬀects (see Fig. A21).

4.3. Macroeconomic uncertainty

We next address macroeconomic uncertainty. All results shown thus

far are generated by WEGDYN, a CGE model that assumes the optimal

macroeconomic state of full capacity utilization (or constrained resources

in production). Additionally, we here deploy E3ME, a PKME model that

assumes idle available resources. Again, we investigate the main scenario

(i.e. SSP2, globally reluctant climate policy, full capacity) and now

compare results to those obtained from E3ME (SSP2, globally reluctant

climate policy, idle available resources).

To understand the diﬀerences in macroeconomic eﬀects between

WEGDYN and E3ME, we need to understand for each model how the

technology cost changes translate into price changes and how these

further pass through the economic system. As explained in Section 4.1,

in WEGDYN changes in unit costs are treated as changes in sectoral

productivity. Higher/lower unit costs, means that more/less inputs are

needed to create one unit of output. This change in productivity im-

mediately translates into relative price changes (which perfectly and

immediately pass through the economic system) and GDP changes.

In E3ME the key factor for price setting is also unit costs, however

an important diﬀerence with WEGDYN is that economic structures are

more rigid. Intermediate production structures (excluding energy in-

puts) have ﬁxed input coeﬃcients. This means that sectors, which need

steel in their production as an input, cannot substitute steel by other

inputs.

Put diﬀerently, iron and steel inputs to other industries de-

pend on their outputs and do not respond directly to price changes.

-25%

-20%

-15%

-10%

-5%

+5%

+10%

2011

2016

2021

2026

2031

2036

2041

2046

2050

tuptuoleetsdnanoriniegnahC

SSP2 - high cost (main sce nario)

SSP3 - high cost

SSP5 - high cost

SSP2 - low cost (main scen ario)

SSP3 - low cost

SSP5 - low cost

-2. 0%

-1. 5%

-1. 0%

-0. 5%

0.0 %

+0.5%

+1.0%

2011

2016

2021

2026

2031

2036

2041

2046

2050

Change in GDP

SSP2 - high cost (main scenario)

SSP3 - high cost

SSP5 - high cost

SSP2 - low cost (main scenario)

SSP3 - low cost

SSP5 - low cost

-2. 0%

-1. 5%

-1. 0%

-0. 5%

0.0 %

+0.5%

+1.0%

2011

2016

2021

2026

2031

2036

2041

2046

2050

iegnah

glob. reluctant - high cost (main scenario)

glob. ambitious - high cost

EU ambitious - high cost

glob. reluctant - low cost (main sce nario)

glob. ambitious - low cost

EU ambitious - low cost

-2. 0%

-1. 5%

-1. 0%

-0. 5%

0.0 %

+0.5%

+1.0%

2011

2016

2021

2026

2031

2036

2041

2046

2050

Change in GDP

full capacity - high cost (main scenario)

full capacity - low cost (main scenario)

idle available resources - high cost

idle available resources - low cost

(a) Socio-economic uncertainty (b) Socio-economic uncertainty

Fig. 5. (a) Change of EU-wide iron and steel output, relative to the baseline with SSP variations on the socio-economic uncertainty layer (assuming reluctant climate

policy and full capacity utilization [WEGDYN]). (b) Change of EU-wide GDP, relative to the baseline with SSP variations on the socio-economic uncertainty layer

(assuming reluctant climate policy and full capacity utilization [WEGDYN]). (c) Change of EU-wide GDP, relative to the baseline in diﬀerent policy-worlds, assuming

SSP2 and full capacity utilization [WEGDYN] (globally reluctant: €46/tCO

by 2050; globally ambitious: €138/tCO

by 2050; EU ambitious: €138/tCO

in EU-ETS

only and €46/tCO

in the rest of the world). (d) Change of EU-wide GDP, relative to the baseline assuming SSP2, globally reluctant climate policy and variations on

the macroeconomic uncertainty layer (“full capacity utilization”[WEGDYN] or “available resources”[E3ME] assumption). Note that (a) is scaled diﬀerently.

Which is to a limited extent the case in WEGDYN. This is implemented via

nested constant elasticity of substitution functions in the production function of

producers. Elasticities of substitution across non-energy intermediate inputs

range between zero and 0.6 (based on Okagawa and Ban (2008)).

G. Bachner, et al. Ecological Economics 172 (2020) 106631

Thus, in E3ME, pass-through of higher costs along the supply chain is

not necessarily the case. In addition, higher costs are also compensated

via lower proﬁts. Hence, overall indirect eﬀects are limited. Ultimately,

the direct and (very limited) indirect cost changes arrive through the

supply chain at ﬁnal demand in the form of higher consumer prices,

where substitution between ﬁnal consumption categories is possible.

Looking at the price eﬀects, we see that with a high-cost speciﬁca-

tion (Fig. 6, top), eﬀects mostly coincide in terms of direction, magni-

tude and development over time. We see somewhat stronger eﬀects

when assuming full capacity utilization, due to higher capital demand,

capital scarcity and the resulting feedbacks on capital markets (higher

capital prices/rents), whereas there are spare capacities (available re-

sources) in the alternative model. With a low-cost speciﬁcation (Fig. 6,

bottom), the price eﬀects also largely coincide, but only until 2031. In

general, after this point in time additional capital costs for investment

repayments (annuities) start to fall again, accompanied with the in-

creasing carbon price. With a full capacity utilization assumption, unit

costs –and thus prices –are ultimately lower than in the baseline by

then and thus prices are also lower. When assuming available resources

(which reﬂects non-optimal states and rigidities), though, prices do not

react immediately and less strong, and prices do not fall below the

baseline levels by 2050. Initially, the price increase is primarily driven

by increases in steel production costs. After the investment period

(when CAPEX costs are repaid), prices do decrease again, but slower as

compared to the full capacity assumption, due to rigidities. The extent of

the price increase partly depends on the competitive import price, in

which case proﬁts will be lower; in any case, proﬁts are the residual. We

also observe that the price eﬀects are distributed more uniformly across

regions. This eﬀect originates from stronger market integrations across

regions in E3ME

than in the WEDGYN model, even though the main

mechanisms for price formation are similar. Note, that in E3ME Greece

acts largely as a price taker in the European market and thus the price

for steel in Greece is very sensitive to technology changes in the rest of

the EU.

In the Post Keynesian model, it is only ﬁnal demand (consumption)

that can react to price changes, as there are ﬁxed input coeﬃcients in

production. As cost pass-through from production output (e.g. raw

material) to consumption goods is limited, demand does not respond

strongly to price changes of iron and steel. Also, full cost pass-through

to ﬁnal demand only happens in the long term (‘Invariance of Tax

Incidence’-assumption, see Section 3.4.3). The impact of the price in-

creases is to reduce real incomes, rather than the demand for iron and

steel. Consequently output does not change substantially. Changes of

output range between ± 0.5%, which are driven by the investment

stimulus (dominating the limited demand reduction from higher

prices). After the investment phase (starting in 2036) eﬀects are very

small, with slight variations across regions, due to competition eﬀects

from varying import price levels. Under the assumption of full capacity

utilization output reacts immediately, more strongly and in diﬀerent

directions, depending on regional characteristics and the assumed cost

speciﬁcation (see Section 4.1, and Figs. A15 and A22 for regional de-

tails on output).

At the aggregate level (Fig. 5d) and under the available resources

assumption, EU-wide GDP is higher by up to +0.25% during the

transition phase and remains nearly unchanged thereafter (for both

technology cost cases). This clearly indicates that the GDP dynamics are

dominated by the short-run eﬀects, i.e. investment requirements in the

iron and steel sector. This is because capital accumulation and capital

productivity does not change via the changing unit costs and prices, as

the economy is not at full capacity in the baseline and compensates

potential productivity losses by activating available resources.In

2011

2021

2031

2041

2050

-6%

-4%

-2%

+2%

+4%

+6%

+8%

+10%

+12%

pstekra

nie

AUT EEU NEU SEU WEU GRC

high cost

2011

2021

2031

2041

2050

2011

2021

2031

2041

2050

2011

2021

2031

2041

2050

2011

2021

2031

2041

2050

-6%

-4%

-2%

+2%

+4%

+6%

+8%

+10%

+12%

2011

2021

2031

2041

2050

eci

pstekramnie

low cost

full capa city (main scenario)

idle ava ilable resou rces

Fig. 6. Change of regional iron and steel market prices, relative to the baseline assuming SSP2, globally reluctant climate policy and variations on the macroeconomic

uncertainty layer (“full capacity utilization”[WEGDYN] or “available resources”[E3ME]).

As parameter settings as well as elasticities of substitution play an im-

portant role in both models, we provide more details in Appendices A2 and A3.

For WEGDYN the interested reader can also ﬁnd all the nested production and

consumption functions.

In E3ME there is more homogeneity between domestically produced goods

and imported goods within the EU, than in WEGDYN, where there is stronger

product diﬀerentiation according to Armington (1969).

G. Bachner, et al. Ecological Economics 172 (2020) 106631

contrast, under the full capacity utilization assumption it is long-run

productivity that dominates GDP eﬀects. Thus, we observe that un-

certainties are large with regard to the macroeconomic state (see Figs.

A24 and A25 regional GDP and welfare eﬀects).

4.4. Discussion

Our results are condensed in Fig. 7. The solid bars (grey) show the

range of long-run eﬀects, i.e. eﬀects in 2050, and how these eﬀects

might change, when changing assumptions on diﬀerent uncertainty

layers. As a point of reference we indicate the results from the main

scenario as a horizontal dashed line (red, i.e. SSP2, globally reluctant

climate policy and full capacity utilization [WEGDYN]). In addition to

the long-run eﬀect we also show short-run eﬀects: The hatched bars

(blue) extend the bars of the long-term eﬀects (grey), thereby indicating

the total range of eﬀects when also taking into account all the

years < 2050; i.e. across time.

Hence, the short-run bars simply ex-

tend the uncertainty ranges that are given from the long-run perspec-

tive to visualize that eﬀects might be stronger or weaker along the way

until 2050 than in 2050. Moreover, we show the variation of regional

eﬀects (within the main scenario), originating from diﬀerences in

-2.5%

-2.0%

-1.5%

-1.0%

-0.5%

0.0%

+0 .5%

+1 .0%

+1 .5%

GDP

socio-economic

climate po licy

macro-economic

regional*

-2.5%

-2.0%

-1.5%

-1.0%

-0.5%

0.0%

+0 .5%

+1 .0%

+1 .5%

socio-economic

climate po licy

macro-economic

regional*

afleW

"hig h cost"

(mature)

"lo w co st"

(infant)

Iron and Steel technology

EU i n 2050, l ong- run (mai n scenari o)

long -r un rang e (i n 2050)

shor t- r un extensi on (< 2050)

Fig. 7. Uncertainty ranges of EU-wide GDP and welfare change, relative to baseline, across scenario and model results (main scenario result levels are indicated by

red dashed line). Left/right columns give high/low technology cost speciﬁcation. (* = Excluding Greece, as no technology switch is implemented in this region.) (For

interpretation of the references to color in this ﬁgure legend, the reader is referred to the web version of this article.)

Only taking into account years in which production of the new technolo-

gies is already running (starting in 2036).

G. Bachner, et al. Ecological Economics 172 (2020) 106631

regional characteristics (indicated by the columns labelled “regional”).

Finally, the diﬀerence between left and right plots shows the techno-

logical uncertainty (high versus low-cost technology speciﬁcation).

Starting the discussion at the technological uncertainty we ﬁnd that

for the high-cost technology speciﬁcation in the short-run eﬀects are

less severe than in the long-run (i.e. by 2050). By contrast, with a low-

cost technology speciﬁcation long-run eﬀects turn out to be beneﬁcial,

but at the cost of higher negative impacts in the short-term. Note, that

our analysis assumes a strict technology standards, which private steel

companies have to comply with. The choice of a production technology

that meets this standard is, however, made by business executives (but

is unknown) and depends upon various factors besides (expected)

technology costs. From such a perspective, it might be too risky for

ﬁrms to choose a technology which is still in its infancy, although it is

expected to be the low-cost alternative in the long-run, with potential

sectoral and macroeconomic gains. If there is a general policy that in-

centivizes low-carbon technology adoptions, managers may still opt for

the less risky and more mature, but in the long-term costlier tech-

nology, and society would have to accept more signiﬁcant macro-losses.

Regarding socio-economic uncertainty, we ﬁnd that the eﬀects are

relatively robust; however, socio-economic uncertainties also concern

regional diﬀerences, such as capital and labor intensities, the degree to

which regions rely on steel imports, or the regional and inter-sectoral

dependency on iron and steel itself. We ﬁnd that the macroeconomic

eﬀects are co-determined relatively strongly by such regional char-

acteristics. In addition, over the course of economic development, dif-

ferent regions might show diﬀerent trajectories regarding e.g. growth of

the labor force and capital stock as well as the development of relative

prices, which co-determine the regional eﬀects. We thus ﬁnd that there

is a relatively broad regional spread. For the high-cost speciﬁcation we

observe GDP losses of up to of −2.3% (Eastern EU), whereas the “lucky

loser”is confronted with only −1.3% (Northern EU). To isolate the

most important regional drivers of uncertainty, as we have indicated

here, is an important, new and promising avenue for future research.

Uncertainty originating from the climate policy layer seems to be

smaller than from other layers, as we observe a clear signal in direction

of results. Not surprisingly, we ﬁnd that in a world with a more strin-

gent climate policy, the potential losses from a transition are smaller

and potential gains are larger. Interestingly, the low-cost technology

speciﬁcation shows larger uncertainties at the climate policy layer,

since the carbon price can trigger a stronger relative competitive ad-

vantage over the conventional technology. Thus, a globally ambitious

climate policy becomes the least detrimental for GDP and welfare. We

also show that EU-wide GDP and welfare eﬀects are rather insensitive

to non-EU policy, however this result is driven by the standard model

setup itself (foreign trade closure of ﬂexible exchange rate and ﬁxed

current account balance). We thus identify foreign trade sensitivity as

another source of uncertainty, which should be addressed in future

research.

A signiﬁcant result is that there is a considerable uncertainty range

from the macroeconomic layer. For the high-cost technology speciﬁ-

cation EU-wide GDP eﬀects range from −1.6% to close to zero. The

possible negative impact results from the full capacity utilization as-

sumption, whereas the less severe impact (close to zero) results from

assuming (idle) available resources. For the latter case, we even observe

a temporary positive eﬀect in the short-run, due to the investment sti-

mulus. For the low-cost technology speciﬁcation, we ﬁnd that the short-

run eﬀects for the full capacity case might be negative; however, long-

run eﬀects are positive throughout. Again, there are large uncertainties

on the macroeconomic layer, with eﬀects close to zero when assuming

available resources.

Other potential sources of uncertainty are climate change impacts

for the economy as a whole and for the iron and steel sector itself. As

soon as the steel sector switches from coke-based production to re-

newable hydrogen-based production, the risks of being adversely af-

fected by climate policy might be lower. However, existing impact

studies (e.g. Solaun and Cerdá (2019) oﬀer a comprehensive review)

indicate increasing risk from physical climate change impacts, poten-

tially aﬀecting also renewable electricity (hydrogen) generation. Hence,

new technologies might be competitive, but nevertheless business ex-

ecutives might be reluctant to switch due to a deteriorating un-

certainty/risk proﬁle in the value chain. An integrated assessment of

climate change impacts and climate mitigation therefore becomes es-

sential.

Regarding macroeconomic uncertainty we explained the funda-

mental diﬀerences in concepts and methodology between the two ap-

plied macroeconomic models (neoclassical CGE models and PKME) in

Section 3.4.1.

Many of the diﬀerences that are relevant here (as-

sumptions on rationality, optimality, money neutrality, labor market

dynamics) are reﬂected in the assumption between full capacity utili-

zation and available resources. However, in addition to these general

diﬀerences there are also model-speciﬁcdiﬀerences between WEGDYN

(the CGE used here) and E3ME (the PKME used here). These diﬀerences

are determined by the modelers rather than by the theoretical under-

pinnings of the models and might inﬂuence diﬀerences in results. We

argue that model-speciﬁcdiﬀerences add to the translational un-

certainty at the science-policy interface and need attention when dis-

cussing our results. We have grouped these diﬀerences into three sets.

First, there are diﬀerences for the labor market. Both models assume

market imperfection in the labor market; WEGDYN in the sense of

classical unemployment, set by a minimum wage that is bound to the

consumer price index of the baseline scenario. E3ME includes un-

employment (voluntary or otherwise) as deﬁned in oﬃcial data as a

regional/country variable in the model. The unemployment rate aﬀects

consumers' expenditure and investment in dwellings by region and

wage rates by industry and region. Second, there are diﬀerences re-

garding the relationship between savings and investment. In WEGDYN

savings drive investments, according to a constant savings rate. New

investments (compared to those in the baseline) are assumed to crowd

out consumption. In E3ME, new investments are assumed to be funded

by banks creating new debt, without any crowding out.

Third, there

are diﬀerences regarding foreign trade. WEGDYN assumes a ﬁxed cur-

rent account balance (foreign savings) with ﬂexibility in exchange rate

and trade patterns. In E3ME, exchange rates are exogenous. Current

account balance of payments deﬁcits or surpluses are assumed to be

matched by capital account ﬂows. These diﬀerences require a more in-

depth and systematic comparison to further disentangle uncertainties at

the macroeconomic layer, i.e. fundamental macroeconomic diﬀerences

and model-speciﬁcdiﬀerences, which should help in reducing transla-

tional uncertainty.

To summarize, here we show that methodological uncertainties as

well as uncertainties at the science-policy interface are large and give

ﬁrst quantiﬁed estimates on aggregate. Hence, we do not claim com-

pleteness in our analysis, but give ﬁrst structured insights into un-

certainties from running diﬀerent models in their standard setups. This

approach should raise awareness that results might strongly depend on

which modelling groups are hired for policy analysis (as typically

modelling groups use their models in the standard setup). However,

results should be compared carefully, as we cannot claim a ceteris

paribus assumption. Still, we demonstrate the importance of carefully

choosing models.

The interested reader will ﬁnd further information and good summaries of

each school in the Exploring Economics website https://www.exploring-

economics.org/en/.

Note that this diﬀerence actually relates to the full capacity utilization

versus available resources distinction. In WEGDYN capital is scarce, as the

economy runs at full capacity, whereas in E3ME money is created by banks,

with new ﬁnance only limited by the banks' aversion to risk.

G. Bachner, et al. Ecological Economics 172 (2020) 106631

5. Conclusions

In the literature, various attempts have been made to address un-

certainty within macroeconomic frameworks, however, studies typi-

cally do not go beyond technical uncertainty, and therefore do not

address methodological and epistemological uncertainty. For the ex-

ample of the European iron and steel industry's transition towards

process emission-free production, we explicitly capture for the ﬁrst time

all of these types of uncertainty by carrying out a systematic analysis of

the potential sectoral and economy-wide eﬀects. Additionally, un-

certainties at the science-policy interface, such as translational un-

certainty, are usually ignored, which we discuss as well in this con-

tribution. We show that uncertainties are large, however, they are

diﬀerent in their meaning and should thus be interpreted with care.

For the range of technological uncertainty, we ﬁnd for our example

that macroeconomic implications can be either positive or negative,

depending on the relative technology costs. This implies quite a sig-

niﬁcant uncertainty, which however can be managed by business ex-

ecutives. We conclude that a discussion is needed on how policy can

help to reduce such risks (potential macroeconomic losses) and possibly

induce a mix of technologies to diversify risk. Possible instruments

might be production standards, subsidies or public investments into

research and development. Further, from the relatively large regional

spreads in results we conclude that low-carbon transition pathways

should be region-speciﬁc and that the transferability of regional case

studies to other regions seems to be quite limited. From the analysis of

macroeconomic uncertainty we conclude that the long-run GDP eﬀects

might be between zero and moderately negative for a high-cost tech-

nology speciﬁcation, or between slightly positive and zero for a low-

cost technology speciﬁcation.

Concerning policy, we clearly see that a CO

price is needed to

support the transition, best if increased over time (in real terms).

Furthermore, we conclude that R&D is needed to support a rapid de-

velopment of low-cost technologies. If policy makers incentivize sectors

to use a currently already mature carbon-free technology (e.g. by new

standards), this might result in (moderate) negative GDP eﬀects.

We draw further policy conclusions, speciﬁcally from the compar-

ison of the two applied macroeconomic models (CGE and PKME). First,

we emphasize that the short- and long-run eﬀects might be very dif-

ferent, with potential negative eﬀects in the transition phase, but with

long-run and sustainable macroeconomic beneﬁts. Inter-generational

equity issues should thus be accounted for.

Second, we conclude that from a policy-makers' perspective the

model choice for policy analysis, and the implicit assumptions of the

macroeconomic state, are highly relevant and should be made with

great care. We demonstrate that the choice of the model inﬂuences the

signs and magnitudes of the macroeconomic impacts. Due to this

translational uncertainty may occur (“results that are incomplete or

conﬂicting so that they can be invoked to support divergent policy positions”;

Kunreuther et al. (2014, p. 178)) and policy recommendations might

diﬀer strongly. We see the danger that results could be consciously

(mis)used to push policy agendas. This implies that the research com-

munity as a whole should be fully transparent, in order to inform policy

makers about the underlying assumptions. Co-design and co-production

can help to foster better communication concerning model choice and

setup. Additionally, the results should be presented in the right context,

i.e. together with the conditions under which they are valid. Thus, re-

garding policy recommendations we need to state (as so often) that “it

depends”; in this case strongly on the state of the economy (i.e. whether

resources are available, which is the case during a recession, or if the

economy runs at full capacity, which is the case during an economic

boom). (Climate) policy makers should thus include economic cycles

into their timing and decisions.

Third, we suggest to increase funding for closer cooperation be-

tween diﬀerent macroeconomic modelling groups that use diﬀerent

approaches; such as the one explored in the research reported in this

article, general equilibrium and Post Keynesian modelers. A critical and

constructive collaboration across groups might trigger (highly needed)

model validation studies as well as the development of new models.

Also, it should be tested whether uncertainties between models with

similar theoretical foundations are smaller than across models with

seemingly quite diﬀerent assumptions. We thus conclude that in the

commissioning of modelling studies uncertainty should get more im-

portant, e.g. by calling for multi-model comparisons across diﬀerent

schools of thought. This could also help to foster cooperation between

modelling camps.

Finally, we emphasize that potential negative macroeconomic ef-

fects of our analysis are moderate. In our “worst-case”scenario EU-wide

absolute GDP/welfare is lower by < 2% in 2050, compared to the

Baseline, which is equivalent to reducing the annual GDP growth rate

by 0.05%-points. This might be an acceptable “price”for helping to

solve a major crisis. One could also interpret the negative welfare eﬀect

as a kind of “forced savings”behavior, necessary to enable a green

transition (cf. Kemp-Benedict, 2018), or, as mandatory insurance

against possible catastrophic climate change (Weitzman, 2009).

Declaration of competing interest

The authors declare that they have no known competing ﬁnancial

interests or personal relationships that could have appeared to inﬂu-

ence the work reported in this paper.

Acknowledgements

This work was supported by the EU Research and Innovation

Framework Program Horizon 2020 (grant number 642260; project

“TRANSrisk”) and by the Austrian Climate and Energy Fund (grant

number B769960; project "EconTrans"). We thank Unnada

Chewpreecha for assistance in E3ME modelling and Keith Williges for

proof-reading. Further, we want to thank two anonymous reviewers,

who gave valuable and constructive feedback on earlier versions of this

paper. All remaining errors are ours.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://

doi.org/10.1016/j.ecolecon.2020.106631.

References

Abler, D.G., Rodríguez, A.G., Shortle, J.S., 1999. Parameter uncertainty in CGE modeling

of the environmental impacts of economic policies. Environ. Resour. Econ. 14, 75–94.

https://doi.org/10.1023/A:1008362712759.

Aguiar, A., Narayanan, B., McDougall, R., 2016. An overview of the GTAP 9 data base.

Journal of Global Economic Analysis 1, 181–208. https://doi.org/10.21642/JGEA.

010103AF.

Aldred, J., 2012. Climate change uncertainty, irreversibility and the precautionary

principle. Cambridge J Econ 36, 1051–1072. https://doi.org/10.1093/cje/bes029.

Armington, P.S., 1969. A theory of demand for products distinguished by place of pro-

duction (Une theorie de la demande de produits diﬀerencies d’apres leur origine)

(Una teoria de la demanda de productos distinguiendolos segun el lugar de produc-

cion). StaﬀPapers - International Monetary Fund 16, 159. https://doi.org/10.2307/

3866403.

Babonneau, F., Haurie, A., Loulou, R., Vielle, M., 2012. Combining stochastic optimiza-

tion and Monte Carlo simulation to deal with uncertainties in climate policy assess-

ment. Environ. Model. Assess. 17, 51–76. https://doi.org/10.1007/s10666-011-

9275-1.

Barker, T., Gardiner, B., 1996. Employment, wage formation and pricing in the European

Union: empirical modelling of environmental tax reform. In: Carraro, C., Siniscalco,

D. (Eds.), Environmental Fiscal Reform and Unemployment, Fondazione Eni Enrico

Mattei (FEEM) Series on Economics, Energy and Environment. Springer Netherlands,

Dordrecht, pp. 229–272. https://doi.org/10.1007/978-94-015-8652-8_9.

Barker, T., Anger, A., Chewpreecha, U., Pollitt, H., 2012. A new economics approach to

modelling policies to achieve global 2020 targets for climate stabilisation. Int. Rev.

Appl. Econ. 26, 205–221. https://doi.org/10.1080/02692171.2011.631901.

Basu, S., 2017. Chapter II - evaluation of Hazard and risk analysis. In: Basu, S. (Ed.), Plant

Hazard Analysis and Safety Instrumentation Systems. Academic Press, pp. 83–167.

https://doi.org/10.1016/B978-0-12-803763-8.00002-9.

G. Bachner, et al. Ecological Economics 172 (2020) 106631

Bednar-Friedl, B., Schinko, T., Steininger, K.W., 2012. The relevance of process emissions

for carbon leakage: a comparison of unilateral climate policy options with and

without border carbon adjustment. In: Energy Economics, The Role of Border Carbon

Adjustment in Unilateral Climate Policy: Results From EMF 29. 34, Supplement 2.

pp. 168–180. https://doi.org/10.1016/j.eneco.2012.08.038.

Bjarnadottir, S., Li, Y., Stewart, M.G., 2019. Chapter nine - climate adaptation for housing

in hurricane regions. In: Bastidas-Arteaga, E., Stewar, M.G. (Eds.), Climate

Adaptation Engineering. Butterworth-Heinemann, pp. 271–299. https://doi.org/10.

1016/B978-0-12-816782-3.00009-7.

Böhringer, C., Balistreri, E.J., Rutherford, T.F., 2012. The role of border carbon adjust-

ment in unilateral climate policy: overview of an Energy Modeling Forum study (EMF

29). Energy Econ. 34, S97–S110. https://doi.org/10.1016/j.eneco.2012.10.003.

Cambridge Econometrics, 2019. E3ME Technical Manual, Version 6.1. Cambridge

Econometrics, Cambridge, UK.

Davidson, P., 1991. Is probability theory relevant for uncertainty? A post Keynesian

perspective. J. Econ. Perspect. 5, 129–143.

Dror, Y., 1970. A policy sciences view of future studies: alternative futures and present

action. Technol. Forecast. Soc. Chang. 2, 3–16. https://doi.org/10.1016/0040-

1625(70)90003-X.

Duchin, F., 1998. Structural Economics: Measuring Change in Technology, Lifestyles, and

the Environment. Island Press, Washington, D.C.

Dutt, A.K., 1984. Stagnation, income distribution and monopoly power. Cambridge J

Econ 8, 25–40. https://doi.org/10.1093/oxfordjournals.cje.a035533.

Edenhofer, O., Knopf, B., Barker, T., Baumstark, L., Bellevrat, E., Chateau, B., Criqui, P.,

Isaac, M., Kitous, A., Kypreos, S., Leimbach, M., Lessmann, K., Magné, B., Scrieciu, Ş.,

Turton, H., Van Vuuren, D.P., 2010. The economics of low stabilization: model

comparison of mitigation strategies and costs. Energy J. 31, 11–48.

Fontana, G., Sawyer, M., 2016. Towards post-Keynesian ecological macroeconomics.

Ecol. Econ. 121, 186–195. https://doi.org/10.1016/j.ecolecon.2015.03.017.

Fujimori, S., Kainuma, M., Masui, T., Hasegawa, T., Dai, H., 2014. The eﬀectiveness of

energy service demand reduction: a scenario analysis of global climate change mi-

tigation. Energy Policy 75, 379–391. https://doi.org/10.1016/j.enpol.2014.09.015.

Funtowicz, S.O., Ravetz, J.R., 1990. Uncertainty and quality in science for policy. In:

Philosophy and Methodology of the Social Sciences, Philosophy and Methodology of

the Social Sciences. Kluwer Academic Publishers, Dordrecht, The Netherlands.

Georgescu-Roegen, N., 1954. Choice, expectations and measurability. Q. J. Econ. 68,

503–534. https://doi.org/10.2307/1881875.

Gowdy, J., Erickson, J.D., 2005. The approach of ecological economics. Cambridge J Econ

29, 207–222. https://doi.org/10.1093/cje/bei033.

Guivarch, C., Rozenberg, J., Schweizer, V., 2016. The diversity of socio-economic path-

ways and CO

emissions scenarios: insights from the investigation of a scenarios

database. Environ. Model Softw. 80, 336–353. https://doi.org/10.1016/j.envsoft.

2016.03.006.

IEA, 2016. World Energy Outlook 2016. International Energy Agency.

Interagency Working Group on Social Cost of Carbon (IWGSCC), 2015. Technical Support

Document: Technical Update of the Social Cost of Carbon for Regulatory Impact

Analysis Under Executive Order 12866. US Government, Washington, DC.

Jansen, H., Klaassen, G., 2000. Economic impacts of the 1997 EU energy tax: simulations

with three EU-wide models. Environ. Resour. Econ. 15, 179–197.

Kemp-Benedict, E., 2018. Investing in a green transition. Ecol. Econ. 153, 218–236.

https://doi.org/10.1016/j.ecolecon.2018.07.012.

Keynes, J.M., 1921. A Treatise on Probability. Macmillan, London.

Keynes, J.M., 1937. The general theory of employment. Q. J. Econ. 51, 209–223. https://

doi.org/10.2307/1882087.

Knight, F., 1921. Risk, Uncertainty, and Proﬁt. Houghton Miﬄin Company, Boston.

Kober, T., Summerton, P., Pollitt, H., Chewpreecha, U., Ren, X., Wills, W., Octaviano, C.,

McFarland, J., Beach, R., Cai, Y., Calderon, S., Fisher-Vanden, K., Rodriguez, A.M.L.,

2016. Macroeconomic impacts of climate change mitigation in Latin America: a cross-

model comparison. Energy Econ. 56, 625–636. https://doi.org/10.1016/j.eneco.

2016.02.002.

Kunreuther, H., Gupta, S., Bosetti, V., Cooke, R., Dutt, V., Ha-Duong, M., Held, H., Llanes-

Regueiro, J., Patt, A., Shittu, E., 2014. Integrated risk and uncertainty assessment of

climate change response policies. In: Edenhofer, O., Pichs-Madruga, R., Sokona, Y.,

Farahani, E., Kadner, S., Seyboth, K., Adler, A., Baum, I., Brunner, S., Eickemeier, P.,

Kriemann, B., Savolainen, J., Schlömer, S., von Stechow, C., Zwickel, T., Minx, J.C.

(Eds.), Climate Change 2014: Mitigation of Climate Change. Contribution of Working

Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate

Change. Cambridge University Press, Cambridge, United Kingdom and New York,

NY, USA.

Lucarelli, B., 2010. Money and Keynesian Uncertainty (MPRA Paper No. 28862).

(Sydney).

Machina, M.J., 1987. Choice under uncertainty: problems solved and unsolved. J. Econ.

Perspect. 1, 121–154. https://doi.org/10.1257/jep.1.1.121.

Mayer, J., Bachner, G., Steininger, K.W, 2019... Macroeconomic implications of switching

to process-emission-free iron and steel production in Europe. Journal of Cleaner

Production 210, 1517–1533. https://doi.org/10.1016/j.jclepro.2018.11.118.

Mayer, J., Bachner, G., Steininger, K.W., 2019. Macroeconomic implications of switching

to process-emission-free iron and steel production in Europe. J. Clean. Prod. 210,

1517–1533. https://doi.org/10.1016/j.jclepro.2018.11.118.

Mercure, J.-F., Knobloch, F., Pollitt, H., Paroussos, L., Scrieciu, S.S., Lewney, R., 2019.

Modelling innovation and the macroeconomics of low-carbon transitions: theory,

perspectives and practical use. Clim. Pol. 19, 1019–1037. https://doi.org/10.1080/

14693062.2019.1617665.

Meyer, B., Ahlert, G., 2019. Imperfect markets and the properties of macro-economic-

environmental models as tools for policy evaluation. Ecological Economics, Resource

Eﬃciency: Concepts, Challenges, Scenarios and Policy Options 155, 80–87. https://

doi.org/10.1016/j.ecolecon.2017.06.017.

Morales-Torres, A., Escuder-Bueno, I., Serrano-Lombillo, A., Castillo Rodríguez, J.T.,

2019. Dealing with epistemic uncertainty in risk-informed decision making for dam

safety management. Reliability Engineering & System Safety 191, 106562. https://

doi.org/10.1016/j.ress.2019.106562.

O’Connell, J., 1996. Uncertainty in “A Treatise on Probability”and the “General Theory”

(Working Paper), Working Paper Series. National University of Ireland, Galway,

Galway, Ireland.

Okagawa, A., Ban, K., 2008. Estimation of Substitution Elasticities for CGE Models.

O’Neill, B.C., Kriegler, E., Riahi, K., Ebi, K.L., Hallegatte, S., Carter, T.R., Mathur, R.,

Vuuren, D.P. van, 2014. A new scenario framework for climate change research: the

concept of shared socioeconomic pathways. Clim. Chang. 122, 387–400. https://doi.

org/10.1007/s10584-013-0905-2.

O’Neill, B.C., Kriegler, E., Ebi, K.L., Kemp-Benedict, E., Riahi, K., Rothman, D.S., van

Ruijven, B.J., van Vuuren, D.P., Birkmann, J., Kok, K., Levy, M., Solecki, W., 2017.

The roads ahead: narratives for shared socioeconomic pathways describing world

futures in the 21st century. Glob. Environ. Chang. 42, 169–180. https://doi.org/10.

1016/j.gloenvcha.2015.01.004.

Pang, J., Shi, Y.-C., Feng, X.-Z., Wu, S.-Y., Sun, W.-L., 2014. Analysis on impacts and co-

abatement eﬀects of implementing the low carbon cement standard. Adv. Clim.

Chang. Res. 5, 41–50. https://doi.org/10.3724/SP.J.1248.2014.041.

Pollitt, H., Mercure, J.-F., 2018. The role of money and the ﬁnancial sector in energy-

economy models used for assessing climate and energy policy. Clim. Pol. 18,

184–197. https://doi.org/10.1080/14693062.2016.1277685.

Refsgaard, J.C., van der Sluijs, J.P., Højberg, A.L., Vanrolleghem, P.A., 2007. Uncertainty

in the environmental modelling process –a framework and guidance. Environ. Model

Softw. 22, 1543–1556. https://doi.org/10.1016/j.envsoft.2007.02.004.

Reilly, M., Willenbockel, D., 2010. Managing uncertainty: a review of food system sce-

nario analysis and modelling. Philos. Trans. R. Soc. Lond. Ser. B Biol. Sci. 365,

3049–3063. https://doi.org/10.1098/rstb.2010.0141.

Rezai, A., Stagl, S., 2016. Ecological macroeconomics: introduction and review. Ecol.

Econ. 121, 181–185. https://doi.org/10.1016/j.ecolecon.2015.12.003.

Rockström, J., Gaﬀney, O., Rogelj, J., Meinshausen, M., Nakicenovic, N., Schellnhuber,

H.J., 2017. A roadmap for rapid decarbonization. Science 355, 1269–1271. https://

doi.org/10.1126/science.aah3443.

Roelich, K., Giesekam, J., 2019. Decision making under uncertainty in climate change

mitigation: introducing multiple actor motivations, agency and inﬂuence. Clim. Pol.

19, 175–188. https://doi.org/10.1080/14693062.2018.1479238.

Runde, J., 1990. Keynesian uncertainty and the weight of arguments. Econ. Philos. 6,

275–292. https://doi.org/10.1017/S0266267100001255.

Sabat, K.C., Murphy, A.B., 2017. Hydrogen plasma processing of iron ore. Metall. Mater.

Trans. B Process Metall. Mater. Process. Sci. 48, 1561–1594. https://doi.org/10.

1007/s11663-017-0957-1.

Schinko, T., Bednar-Friedl, B., Steininger, K.W., Grossmann, W.D., 2014. Switching to

carbon-free production processes: implications for carbon leakage and border carbon

adjustment. Energy Policy 67, 818–831.

Shackle, G.L., 1943. The expectational dynamics of the individual. Economica 10,

99–129. https://doi.org/10.2307/2549459.

Solaun, K., Cerdá, E., 2019. Climate change impacts on renewable energy generation. A

review of quantitative projections. Renew. Sust. Energ. Rev. 116, 109415. https://

doi.org/10.1016/j.rser.2019.109415.

Sommer, M., Kratena, K., 2017. The carbon footprint of European households and income

distribution. Ecol. Econ. 136, 62–72. https://doi.org/10.1016/j.ecolecon.2016.12.

008.

UNFCCC, 2017. Greenhouse Gas Inventory Data. United Nations.

Walker, W.E., Harremoës, P., Rotmans, J., Sluijs, J.P. van der, Asselt, M.B.A. van, Janssen,

P., Krauss, M.P.K. von, 2003. Deﬁning uncertainty: a conceptual basis for uncertainty

management in model-based decision support. Integr. Assess. 4, 5–17. https://doi.

org/10.1076/iaij.4.1.5.16466.

Wang, C., Chen, J., 2006. Parameter uncertainty in CGE modeling of the macroeconomic

impact of carbon reduction in China. Tsinghua Sci. Technol. 11, 617–624. https://

doi.org/10.1016/S1007-0214(06)70242-5.

Weitzman, M.L., 2009. On modeling and interpreting the economics of catastrophic cli-

mate change. Rev. Econ. Stat. 91, 1–19.

Zappia, C., 2014. Non-Bayesian decision theory ahead of its time: the case of G. L. S.

Shackle. Camb. J. Econ. 38, 1133–1154. https://doi.org/10.1093/cje/beu023.

G. Bachner, et al. Ecological Economics 172 (2020) 106631

Are rural households hit hardest? Exploring the distributional effects of region-specific compensation payments in the Austrian CO2 pricing scheme

Article

Full-text available

Dec 2024
ENERG ECON

A systematic review of decarbonization pathway and modeling conception in iron and steel industry at micro-, meso-, and macro-levels

Article

Full-text available

Oct 2024
ENVIRON SCI POLLUT R

To meet the carbon–neutral goal in 2050, an accelerated transition of decarbonization is needed in the iron and steel industry. Lots of decarbonization path models are recently being used to analyze the technology pathways, industrial structure adjustments, and short- or long-term policies for greenhouse gas emission reduction. Thus, systematic insight is required for better making sense of the status quo and differences in low-carbon transition pathways at different levels. This paper reviews the carbon emission reduction of steel production routes and summarizes the decarbonization models from micro-, meso-, and macro-level perspectives, respectively. First, we systematically analyze the capacity and potential of both available and emerging technologies in the iron and steel production process. Additionally, we conduct a theoretical analysis of the bottom-up models currently used as effective tools to assess decarbonization technology pathways. Subsequently, pathways and models in terms of industrial synergy are elaborated. Policy guidance and market support are explored to overcome the challenges of collaborative carbon reduction faced by global steel producers. The characteristics of top-down models for supporting carbon reduction are also discussed. Finally, gaps in the literature and future research agendas are identified. Advancing this research could enrich discussions on steel industry decarbonization and provide clearer assessments of modeling approaches. The results indicate that existing energy-saving technologies in the iron and steel industry have limited carbon reduction capacity. Significant reduction requires coordinated efforts with upstream and downstream industries, particularly in hydrogen and power sectors, along with financial support and favorable policies.

Impact of Macroeconomic Variables on the Economic Growth in the Middle East Countries

Article

Sep 2020

Raed Abueid

This study investigated the impact of macroeconomic variables on the economic growth in the Middle East countries. The objectives of the research were to determine the relationship among GDP, inflation, oil and gas prices as well as competition on the economic growth in the Middle East. The study spanned between 2015 and 2019. The time series data for the region was obtained from the World Bank while data on ROA, together with ROE, was mined on the OPIC report. The GDP, oil price, interest rate, inflation, and completion were collected from OPIC and International monetary fund (IMF) Middle East report. The research used a panel regression method to analyses the impact of macroeconomic variables on the economic growth of Middle East countries. The results showed that the GDP, oil and gases prices, inflation as well as supply of money are the most significant macroeconomic variables of the growth of the economy. The government, together with authorities that regulate, is required to discover other resources apart from hydrocarbon to refresh its economy as well as engender cost-effective financial structure in view of the global low-carbon technology advancements and decarburization policies. The research has a useful, practical implication for the policymakers, regulators of the region. The government, together with authorities that regulate, is required to discover other resources apart from hydrocarbon to refresh its economy as well as engender cost-effective financial structure.

Do mineral imports increase in response to decarbonization indicators other than renewable energy?

Article

Jan 2024
J CLEAN PROD

While previous studies examine the impact of renewable energy generation capacity on mineral imports where minerals are broadly defined, this study examines the role of three decarbonization indicators (energy efficiency, energy's carbon footprint, and renewable generation) in driving import demand for four specific minerals used in multiple decarbonization technologies: copper, aluminum, nickel, and manganese. Our results, based on the cross-sectional autoregressive distributed lag modeling approach applied to panel data from 33 countries, suggest that imports for some minerals increase in response to (1) investments in energy efficiency (aluminum and manganese), (2) reducing energy's carbon footprint (copper), and (3) investments in renewable energy generation (copper and manganese). Nickel imports were found to increase with higher levels of energy use per GDP suggesting their role in cleaner vehicles, which heavily rely on electrification, higher electricity use, and urbanization. When imports are broadly defined to include all types of mineral groups, import values were found to be responsive to higher renewable generation but not the other decarbonization indicators. Our findings provide insights into developing mineral-specific strategies and policy frameworks to ensure supply chain resilience.

Macroeconomic Strategies and Policies Impacting the Decarburization of Economies in the Member Countries of the European Union

Article

Full-text available

Oct 2023

In the context of the strategic development objectives of the European Union for the next strategic cycle until 2030, a priority issue is the decarburization of the economy, oriented towards the design and monitoring of strategies in order to achieve the objectives regarding climate change, the simultaneous stimulation of economic growth and ensuring social cohesion. A special place belongs to the progress of climate change, which requires progressive steps in the implementation of strategies and policies for all member countries of the EU. We propose in this article to focus on the application of decarburization strategies and policies in the countries of the European Union with an emphasis on Romania, with a necessary approach to social inclusion issues. The development of proposals for efficient strategies for the decarburization of economies in EU, the comparative analysis between the diversification of macroeconomic policies and instruments with different evaluation criteria, having as an essential support - the digitalization tools, which will accelerate the process of decarburization of the countries' economy EU. We propose the development of decarburization strategies based on a mix of policies focused on establishing emission prices; standards and regulations; the implementation of capital, labor and innovation reallocation policies. Key words: Decarburization of the economy, strategies and policies, digitalization, innovation

Feasible low-carbon technological pathway: Sustainable development strategies in the vanadium titanium steel industry

Article

Feb 2025
APPL ENERG

Factor reallocation path for low-carbon transformation: A perspective of manufacturing industry ecosystem

Article

May 2024
ENERG ECON

Economy-wide impacts of socio-politically driven net-zero energy systems in europe

Article

Full-text available

Mar 2024
ENERGY

A hybrid electricity-economy model to assess the aggregate impacts of low-carbon transition: An application to France

Article

Nov 2023
ECOL ECON

The article investigates different macroeconomic channels of transmission of the low-carbon transition in two official scenarios for the French power system under different oil price scenarios. A hybrid model combines, with a two-way soft link, a dynamic general equilibrium model and a partial equilibrium model for the power generation mix. The results show that the decarbonization of the economy and growth can be driven by technical progress and substitution mechanisms triggered by low-carbon transition policies. The main transmission channels are the demand for energy through the relative price of energy and the relative efficiency of energy, and the demand for durable goods. Considering the evolution of relative electricity prices, for an economy transitioning toward mass electrification, seems to be key in our model. And, if recycled, the carbon tax may not significantly harm the economic welfare of households but weigh (slightly) on the firms in the long run.

The aggregate effects of the structure of information in low-carbon transition policies: An application to France

Article

Aug 2023

Climate change impacts on renewable energy generation. A review of quantitative projections

Article

Full-text available

Dec 2019
RENEW SUST ENERG REV

Research on climate change impacts on renewable energy is becoming increasingly relevant due to the vulnerability of the sector and to the continual development of methodologies and availability of data. Public and private decision-making needs specific research. However, many gaps still exist in certain geographical regions and technologies. Providing economic estimates with a value chain perspective are also missing from most papers. This paper addresses the most relevant studies that project quantitative estimates of climate change impacts on solar, wind, hydro and other renewable generation technologies. Summary tables of impacts and projections are provided so that researchers, governments and the private sector may have an accurate view of the state-of-the-art on this topic.

Modelling innovation and the macroeconomics of low-carbon transitions: theory, perspectives and practical use

Article

Full-text available

Jun 2019

Energy and climate policies may have significant economy-wide impacts, which are regularly assessed based on quantitative energy-environment-economy models. These tend to vary in their conclusions on the scale and direction of the likely macroeconomic impacts of a low-carbon transition. This paper traces the characteristic discrepancies in models’ outcomes to their origins in different macro-economic theories, most importantly their treatment of technological innovation and finance. We comprehensively analyse the relevant branches of macro-innovation theory and group them into two classes: ‘Equilibrium’ and ‘Non-equilibrium’. While both approaches are rigorous and self-consistent, they frequently yield opposite conclusions for the economic impacts of low-carbon policies. We show that model outcomes are mainly determined by their representations of monetary and finance dimensions, and their interactions with investment, innovation and technological change. Improving these in all modelling approaches is crucial for strengthening the evidence base for policy making and gaining a more consistent picture of the macroeconomic impacts of achieving emissions reductions objectives. The paper contributes towards the ongoing effort of enhancing the transparency and understanding of sophisticated model mechanisms applied to energy and climate policy analysis. It helps tackle the overall ‘black box’ critique, much-cited in policy circles and elsewhere. Key policy insights • Quantitative models commissioned by policy-makers to assess the macroeconomic impacts of climate policy generate contradictory outcomes and interpretations. • The source of the differences in model outcomes originates primarily from assumptions on the workings of the financial sector and the nature of money, and of how these interact with processes of low-carbon energy innovation and technological change. • Representations of innovation and technological change are incomplete in energy-economy-environment models, leading to limitations in the assessment of the impacts of climate-related policies. • All modelling studies should state clearly their underpinning theoretical school and their treatment of finance and innovation. • A strong recommendation is given for modellers of energy-economy systems to improve their representations of money and finance.

Macroeconomic implications of switching to process-emission-free iron and steel production in Europe

Article

Full-text available

Nov 2018
J CLEAN PROD

Climate change is one of the most serious threats to the human habitat. The required structural change to limit anthropogenic forcing is expected to fundamentally change daily social and economic life. The production of iron and steel is a special case of economic activities since it is not only associated with combustion but particularly with process emissions of greenhouse gases which have to be dealt with likewise. Traditional mitigation options of the sector like efficiency measures, substitution with less emission-intensive materials, or scrap-based production are bounded and thus insufficient for rapid decarbonization necessary for complying with long-term climate policy targets. Iron and steel products are basic materials at the core of modern socio-economic systems, additionally being essential also for other mitigation options like hydro and wind power. Therefore, a system-wide assessment of recent technological developments enabling almost complete decarbonization of the sector is substantially relevant. Deploying a recursive-dynamic multi-region multi-sector computable general equilibrium approach, we investigate switches from coke-to hydrogen-based iron and steel technologies in a scenario framework where industry decisions (technological choice and timing) and climate policies are misaligned. Overall, we find that the costs of industry transition are moderate, but still ones that may represent a barrier for implementation because the generation deciding on low-carbon technologies and bearing (macro)economic costs might not be the generation benefitting from it. Our macroeconomic assessment further indicates that anticipated bottom-up estimates of required additional domestic renewable electricity tend to be overestimated. Relative price changes in the economy induce electricity substitution effects and trigger increased electricity imports. Sectoral carbon leakage is an imminent risk and calls for aligned course of action of private and public actors.

Decision making under uncertainty in climate change mitigation: introducing multiple actor motivations, agency and influence

Article

Full-text available

May 2018

Climate change mitigation has two main characteristics that interact to make it an extremely demanding challenge of governance: the complexity of the socio-technical systems that must be transformed to avoid climate change and the presence of profound uncertainties. A number of tools and approaches exist, which aim to help manage these challenges and support long-term decision making. However, most tools and approaches assume that there is one decision maker with clearly defined objectives. The interaction between decision makers with differing perspectives and agency is an additional uncertainty that is rarely addressed, despite the wide recognition that action is required at multiple scales and by multiple actors. This article draws inspiration from dynamic adaptive policy pathways to build on current decision support methods, extending analysis to include the perspectives and agency of multiple actors through a case study of the UK construction sector. The findings demonstrate the importance of considering alignment between perspectives, agency and potential actions when developing plans; the need for mobilizing and advocacy actions to build momentum for radical change; and the crucial influence of interaction between actors. The decision support approach presented could improve decision making by reflecting the diversity and interaction of actors; identifying short-term actions that connect to long-term goals and keeping future options open. Key policy insights • Multiple actors, with differing motivations, agency and influence, must engage with climate change mitigation, but may not do so, if proposed actions do not align with their motivations or if they do not have agency to undertake specific actions. • Current roadmaps, which assume there is one decision maker with control over a whole system, might overstate how effective proposed actions could be. • Decision making under deep uncertainty needs to account for the motivations and agency of diverse decision makers and the interaction between these decision makers. • This could increase the implementation and effectiveness of mitigation activities.

Hydrogen Plasma Processing of Iron Ore

Article

Full-text available

Jun 2017

Iron is currently produced by carbothermic reduction of oxide ores. This is a multiple-stage process that requires large-scale equipment and high capital investment, and produces large amounts of CO2. An alternative to carbothermic reduction is reduction using a hydrogen plasma, which comprises vibrationally excited molecular, atomic, and ionic states of hydrogen, all of which can reduce iron oxides, even at low temperatures. Besides the thermodynamic and kinetic advantages of a hydrogen plasma, the byproduct of the reaction is water, which does not pose any environmental problems. A review of the theory and practice of iron ore reduction using a hydrogen plasma is presented. The thermodynamic and kinetic aspects are considered, with molecular, atomic and ionic hydrogen considered separately. The importance of vibrationally excited hydrogen molecules in overcoming the activation energy barriers, and in transferring energy to the iron oxide, is emphasized. Both thermal and nonthermal plasmas are considered. The thermophysical properties of hydrogen and argon–hydrogen plasmas are discussed, and their influence on the constriction and flow in the of arc plasmas is considered. The published R&D on hydrogen plasma reduction of iron oxide is reviewed, with both the reduction of molten iron ore and in-flight reduction of iron ore particles being considered. Finally, the technical and economic feasibility of the process are discussed. It is shown that hydrogen plasma processing requires less energy than carbothermic reduction, mainly because pelletization, sintering, and cokemaking are not required. Moreover, the formation of the greenhouse gas CO2 as a byproduct is avoided. In-flight reduction has the potential for a throughput at least equivalent to the blast furnace process. It is concluded that hydrogen plasma reduction of iron ore is a potentially attractive alternative to standard methods.

Climate Adaptation for Housing in Hurricane Regions

Chapter

Jan 2019

Dealing with epistemic uncertainty in risk-informed decision making for dam safety management

Article

Jul 2019
RELIAB ENG SYST SAFE

Adrian Morales-Torres

In recent years, the application of risk analysis to inform dam safety governance has increased significantly. In this framework, considering explicitly and independently both natural and epistemic uncertainty in quantitative risk models allows to understand the sources of uncertainty in risk results and to estimate the effect of actions, tests, and surveys to reduce epistemic uncertainty. In this paper, Indexes of Coincidence are proposed to analyze the effect of epistemic uncertainty in the prioritization of investments based on risk results, which is the key issue in this paper. These indexes allow consideration of the convenience of conducting additional uncertainty reduction actions. These metrics have been applied to the prioritization of risk reduction measures for four concrete gravity dams in Spain. Results allow for a better understanding of how epistemic uncertainty of geotechnical resistance parameters influence risk-informed decision making. The proposed indexes are also useful for probabilistic risk analyses of other civil engineering structures with high epistemic uncertainty environments, since they analyze whether existing uncertainty could have an impact on decision making, outlining the need for extra studies, surveys and tests.

Investing in a Green Transition

Article

Nov 2018
ECOL ECON

Eric Kemp-Benedict

At the Rio + 20 conference the world's governments affirmed a need to transition toward green economies, and background reports in preparation for the event emphasized the role of investment. Subsequent authors noted that private investment flows overwhelm public ones, arguing that private investment must be mobilized for a transition. However, the role of private investment in a major economic transition is poorly understood, particularly when the investment is not inherently attractive to investors and must be made attractive through policy. We present a model representing investment-macroeconomic interactions with classical features. Among its novel features is a distinction between “affected” sectors, which are sensitive to the amount of green capital in the economy because of network effects and forward-backward linkages, and “unaffected” sectors, which are relatively insensitive. We argue that affected sectors are both harder to change and nearer to the base of the economy. In model runs, we observe that in a green transition, the affected sector lags the unaffected sector with possible implications for a transition strategy. Thus, in the model, firm and investor behavior postpones deeper and more systemic changes until late in the transition.

A roadmap for rapid decarbonization

Article

Mar 2017

Although the Paris Agreement's goals (1) are aligned with science (2) and can, in principle, be technically and economically achieved (3), alarming inconsistencies remain between science-based targets and national commitments. Despite progress during the 2016 Marrakech climate negotiations, long-term goals can be trumped by political short-termism. Following the Agreement, which became international law earlier than expected, several countries published mid-century decarbonization strategies, with more due soon. Model-based decarbonization assessments (4) and scenarios often struggle to capture transformative change and the dynamics associated with it: disruption, innovation, and nonlinear change in human behavior. For example, in just 2 years, China's coal use swung from 3.7% growth in 2013 to a decline of 3.7% in 2015 (5). To harness these dynamics and to calibrate for short-term realpolitik, we propose framing the decarbonization challenge in terms of a global decadal roadmap based on a simple heuristic—a “carbon law”—of halving gross anthropogenic carbon-dioxide (CO2) emissions every decade. Complemented by immediately instigated, scalable carbon removal and efforts to ramp down land-use CO2 emissions, this can lead to net-zero emissions around mid-century, a path necessary to limit warming to well below 2°C.

The Carbon Footprint of European Households and Income Distribution

Article

Jun 2017
ECOL ECON

This paper calculates the carbon footprint of private consumption in the EU27 by five groups of household income, using a fully fledged macroeconomic input-output model covering 59 industries and five groups of household income for the EU27. Due to macroeconomic feedback mechanisms, this methodology – besides induced intermediate demand – also quantifies: (i) private consumption induced in the other household groups, (ii) impacts on other endogenous final demand components, and (iii) negative feedback effects due to output price effects of household demand. The carbon footprint is calculated separately for the consumption vector of each of the five income groups. The simulation results yield a non-linear income elasticity of direct and indirect emissions at each income level: the value of the direct footprint income elasticity decreases from 1.32 (first quintile) to 0.69 (fourth quintile). The value of the indirect footprint income elasticity is always below unity and decreases from 0.89 to 0.62. The results in general reveal a relative decoupling effect: the share of the top income group in income (45%) is much larger than its share in the carbon footprint (37%) and vice versa for the bottom income group (6% in income and 8% in footprint).

Uncertainties in macroeconomic assessments of low-carbon transition pathways -The case of the European iron and steel industry

Abstract and Figures

Recommended publications

Macroeconomic implications of switching to process-emission-free iron and steel production in Europe...

Macroeconomic implications of switching to process-emission-free iron and steel production in Europe

Costs or benefits? Assessing the economy-wide effects of the electricity sector's low carbon transit...

Risk assessment of the low-carbon transition of Austria’s steel and electricity sectors