Tom Youngman, Research Project Manager for Macroeconomics of the Seventh Carbon Budget, reflects on the three-day workshop, 'Macroeconomic Implications of Climate-Related Risks: Challenges and Opportunities for the Low-Carbon Transition', held at Scuola Superiore Sant’Anna in Pisa between 23 and 25 February 2026.
His attendance was supported by a travel bursary from Oxford’s ZERO Institute.
Three Pisan Days of Climate Macroeconomic Pluralism
The workshop robustly answered the post-2008 call for pluralism in economics. Economics gatherings tend to focus on one modelling approach, or at best sift different approaches into separate parallel strands. In this workshop the organisers carefully organised each session to contain side-by-side presentations of contrasting methodologies. This made for a slightly increased mental load, switching between mathematical frameworks every 30 minutes, but it fostered the sort of dialogue and co-operation that economics sorely needs.
The workshop was the diffusion event for two EU-funded research projects, using financial support from the Young Scholars Initiative to expand it out to a broader audience and range of presenters. This seems a nice, replicable organising model. Most of the papers I’ll mention here aren’t published yet, but some will be in an upcoming special issue of Economic Modelling.
Three broad categories of climate macroeconomic models
I would categorise the approaches into three broad categories: short run, transition and descriptive. Some might call the first two equilibrium and disequilibrium models, but this feels a poor shorthand given that varieties of equilibrium are present in both kinds of model.
Descriptive
A few papers examined relationships between firms and industrial sectors. Christian Vezil (Universite Cote d’Azur) used network centrality measure PageRank to assess which high-emitting sectors of the economy (as represented in input-output tables) hold the most important positions in the production network. This made the financial sector surge up the rankings in terms of emissions responsibility. Weiwei Bendixen (Queen Mary) found evidence that firms influence each other when deciding whether to participate in the green economy. There seems to be evidence of influence both among geographical neighbours and those in the same industry.
Several papers assessed the impacts of climate policies using econometric methods. Urszula Szczerbowicz (SKEMA Business School) gave a convincing account of how ECB announcements of climate action measures led to asset repricing. The finding only applied to climate action announcements – not general climate intentions or sentiments. Olivier de Bandt (Banque du France) found that climate stress testing by the ECB had led to banks rebalancing their current portfolios, but hadn’t appeared to lead to changes in which new loans they issued. Other novel econometrics papers included Fulvia Marotta (De Nederlandsche Bank) on the impacts of climate policy uncertainty and Elisabetta Cappa (Sant’Anna) on the effect of the EU Emissions Trading Scheme on firms and national macro variables.
Transition
Many of the transition models had a stock-flow consistent (SFC) core.[1][2] This class of model is structured around the UN System of National Accounts, and accounts separately for real flows of goods on the one hand and financial flows of money and other paper or digital assets on the other. These models have been expanded in several directions, from models that disaggregate sectors of the economy using input-output tables, to agent-based models that disaggregate households and firms into heterogeneous digital agents with independent initial conditions. In the ‘transition’ model category I would also place network models and input-output models.
My presentation on how we are using INET Oxford’s macroeconomic model to assess the impacts of the UK’s carbon budgets sat in this category. We presented the approach we are taking and preliminary qualitative findings on the impacts of the six macroeconomic mechanisms we have built into the model. We prepared the presentation jointly between the whole MacroCB7 team, and co-author Pirta Palola also came to the workshop. The paper seemed to be well received, described by discussant Lilit Popoyan (Queen Mary) as ‘analytically clean’ and with potential for policy impact. Her list of things she would like to see closely matched our aspirations for the future: technological learning and diffusion, endogenising emissions and better interrogation of the mechanisms driving our results. We will be taking her suggestion to experiment with how the financing of green subsidies affects their impacts. Francesco Lamperti (Sant’Anna) questioned why our confidence intervals were so wide and Jean François Mercure (Exeter) queried whether the model could represent credit risk and differentiated regional impacts.
There were many other standout agent-based models. Emanuele Ciola (Università di Firenze) presented MATRIX, an agent-based macroeconomic model allowing for migration, something other similar models abstract away from. He used it to assess the impacts of climate change on different European regions. Gianluca Pallante (Sant’Anna) incorporated a sophisticated interbank network into the Dystopian Schumpeter meets Keynes model.[3] Jessica Reale (IUSS Pavia) gave probably the slickest presentation of the workshop, explaining her agent-based SFC model incorporating S-shaped technological diffusion curves into firms’ productivity functions. I’ll definitely be drawing inspiration from this as we seek to draw on Benjamin Wagenvoort and Brendon Tankwa’s technology diffusion research in the MacroCB7 project.
There were sophisticated interdisciplinary models of human-environmental interactions. David An (Università di Firenze) and Jlenia di Noia (IUSS Pavia) both presented models with sophisticated hydrological components. There were also ecological economic models of climate impacts on supply chain networks from Elisa Nobile (IUSS Pavia / IIASA) and Camilla Pelosi (Université Paris 1 Panthéon-Sorbonne). The projects involving interdisciplinary collaborations seemed stronger, but were sometimes harder to interpret as abbreviations for soft-linked environmental models then often replaced actual specification of the functions in the model.
Short run
The short run business cycle models were largely a class of models referred to as Dynamic Stochastic General Equilibrium (DSGE)[4] models, that broadly build upon the Real Business Cycle framework.[5] As someone literate in the framework but never a practitioner, I was interested to see that many cutting-edge variants include not just one or two but many of the modifications proposed to DSGE models since 2008. The ‘Environmental’ DSGE models (E-DSGE) presented included longstanding New Keynesian features like monopolistic competition[6] but also newer propositions like financial intermediation[7] and heterogeneous consumer classes.[8]
The innovations to the DSGE framework presented at the workshop drew from a plurality of schools of thought. Matthias Kaldorf (Deutsche Bundesbank) presented a DSGE model with Minskyan financial instability.[9] He found that the welfare implications of financial instability arising due to the climate transition are an order of magnitude lower than the welfare improvement from tackling climate change. I appreciated his graphical presentation of how the choice of discount rate affected the findings, bringing refreshing transparency to one of the most contested variables in environmental economics. Luca Lochi (Lancaster University) incorporated a Markov-switching model of ‘bull’ and ‘bear’ oil price regimes into an E-DSGE model, nicely incorporating time series approaches from the finance literature.
The sceptic in me asks whether this is the broadening of the DSGE church, or the shoehorning of alternative ideas into a modelling paradigm that poorly accommodates them. In his keynote, Jean François Mercure emphasised how important heavy-tailed distributions and tipping points are to many types of climate models, but also to many classes of economic relationships. These are features of the economic world poorly represented in DSGE models. In a roundtable session between practitioners of all three modelling approaches, environmental DSGE modeller Francesca Diluiso (Bank of England) affirmed that the short run DSGE framework is not the best approach for assessing longer term, structural changes to the economy. Diluiso argued that different economic modelling approaches need to be in closer dialogue with each other, with researchers able to better explain the causes of variation in findings. The dialogue promoted in this workshop was an excellent first step towards that goal.
