Green swans: climate change risks, central banking and financial stability

by Luiz Awazu Pereira da Silva, Bank of International Settlements, drafted by Shashwat Koirala, OECD Economics Department

The Chief Economist Talks are part of the OECD’s high-level distinguished speaker series in which global economic leaders, top thinkers and decision makers are invited to discuss their perspectives on the world economy with the OECD Chief Economist. The talks aim to foster learning and inspiration and provoke meaningful discussions. Previous speakers have included: Claudio Borio (BIS), Peter Praet (ECB), Maurice Obstfeld (IMF), Penny Goldberg (World Bank), Debora Revoltella (EIB), Hal Varian (Google), Sergei Guriev (EBRD), Stefanie Stantcheva (Harvard), Emmanuel Moulin (Ministry of Economy and Finance, France), Philipp Steinberg (Ministry of Economic Affairs and Energy, Germany), and Jean Pisani-Ferry and George Papakonstantinou (EUI). Participation in these events are by invitation only and are aimed at OECD staff and the OECD Ambassadors and delegations. They are not open to the press.

On April 23, 2020, the OECD hosted Luiz Awazu Pereira da Silva, Deputy General Manager, Bank of International Settlements, to discuss his work on the challenges posed by climate change to financial stability, drawing on his co-authored book, “The green swan: Central banking and financial stability in the age of climate change”. This blog presents key takeaways from his talk.

The unprecedented challenge posed by climate change is well documented. The rising concentration of greenhouse gases in the atmosphere has profound environmental impacts (e.g. rising sea levels, extreme temperature events, etc.) that threaten the delicate balance of the planet’s natural systems. The human and societal consequences of the climate emergency are also massive, as environmental damages can exacerbate inequalities, food and water insecurity, and conflicts. Accounting for climate-related risks is, thus, indispensable for building resilient socio-economic-ecological systems.

There is an emerging recognition among central banks and financial regulators that climate-related risks are also a source of price and financial instability, and that there is a need to safeguard the financial system against these risks. This is complicated by a paradoxical tension between physical climate risks and transition risks. For example, on one hand, inaction towards the climate crisis means that climate-related accidents become more frequent and severe, threatening socio-economic systems and financial stability (i.e. physical risks). On the other hand, a rapid and aggressive decarbonisation effort can lead to sudden asset repricing (i.e. transition risks). This tension epitomises the fact that climate-related risks are transmitted through complex and inter-connected channels and have cascade effects. Treating these risks requires a departure from status quo thinking, as outlined by the following four key ideas.

First, while similar in some respects to “black swans”– highly unexpected events with severe far-reaching consequences (e.g. 2008 U.S. housing market crash) that can be best explained ex post – climate-related risks are distinct. They are not tail-risk events; scientific evidence suggests that climate-related shocks are virtually certain to occur, though the exact timing of these events is uncertain. Since the climate crisis poses an existential threat to humanity, climate-related risks are also more catastrophic than traditional systemic financial risks. Finally, as alluded to earlier, climate-related risks are much more complex. They are propagated non-linearly with destructive feedback loops and can cascade across sectors, countries and systems (see Figure 1 for a representation of chain reactions stemming from climate-related risks). Taking inspiration from the “black swan” moniker, climate-related events are termed “green swans”.

Second, a methodological shift in macroeconomic-climate modelling is required to better understand green swan events, and how they emerge, accumulate and cascade. Backward-looking and deterministic approaches (e.g. vector autoregressive models) that extrapolate historical trends do not suffice in capturing the complexity and radical uncertainty of climate-risks. Even current scenario-based forward-looking risks assessment mechanisms are unable to completely incorporate the broad range of chain-reactions associated with climate change. This, in tandem with the fact that these approaches lack granularity and there is uncertainty regarding approaches to climate-change mitigation, means that the current paradigm of models cannot fully elucidate the potential macroeconomic, sectoral and firm-level repercussions of climate change. Thus, an exploration of alternative approaches is needed, such as non-equilibrium models (instead of more sophisticated dynamic stochastic general equilibrium models), sensitivity analysis with more complex scenarios, and studies specific to countries, sectors and firms.

Third, given the intrinsic complexity of climate change, international co-ordination and co-operation is vital. While central banks play a critical role in mitigating climate-related risks, they do not possess a silver bullet to do so by themselves. Central banks and financial regulators have a role to play in identifying and managing climate-risks (e.g. integrating risks into prudential regulation), internalising externalities (e.g. incorporating environmental, social and governance considerations into their own portfolios), and enabling structural low-carbon transitions (e.g. reforming the international monetary and financial system). Nevertheless, many tools, such as green fiscal policy and carbon pricing, fall outside their purview, and uncoordinated actions from central banks would be insufficient and could potentially have unintended consequences. A systems-wide green transition necessitates buy-in and action from all stakeholders (i.e. governments, private sector, and civil society), and central banks need to contribute to coordinate on climate change by being more proactive on this front while continuing to fulfil their financial stability mandate.

Fourth, it is important to acknowledge that green swans have a tremendous negative redistributive impact, within and between countries. Not only do the physical risks stemming from climate change predominately affect lower-income countries, but also the costs of adaptation to climate-change (e.g. shift away from carbon-intensive industries) are higher for poorer households. This means that addressing climate change requires scaled-up mechanisms for redistribution and a redesign of societal safety nets and efforts to finance the green transition of low-income countries. Otherwise, a society-wide acceptance of actions on climate change will prove elusive.

The ecological and environmental stability of the planet is a prerequisite for price and financial stability. So, for central banks to fulfil their central mandate, they have an important role in contributing to a systems-wide climate-change effort. In a nutshell, this involves identifying and communicating the risks ahead, calling for bold actions from all stakeholders to ensure the resilience of the earth’s socio-ecological systems, and helping manage the risks within the bounds of their mandate.


Bolton, P., Despres, M., Pereira da Silva, L. A., Samana, F., & Svartzman, R. (2020). The green swan: Central banking and financial stability in the age of climate change. Bank for International Settlements.

Carbon tax, emissions reduction and employment: Some evidence from France

by Damien Dussaux, Economist, OECD Environment Directorate

In September 2019 the French Parliament adopted the law on energy and climate which enshrines in the French law the objective of Carbon Neutrality by 2050, in line with the 2015 Paris Climate Agreement. Achieving carbon neutrality in France will require a drastic decrease in greenhouse gas (GHG) emissions of 75% by 2050 compared to 1990 levels.

To ensure this target is met, the French government developed a “National Low Carbon Strategy”, which acts as a roadmap for implementing a low-emission transition in each sector of the economy. For example, GHG emissions from industry account for almost one fifth of emissions in France, equivalent to total GHG emissions of Romania, and, under the proposed sectoral plan, will be reduced by a quarter within the next ten years.

France is currently employing two main carbon pricing mechanisms:

  1. European Union Emissions Trading System (EU-ETS), which has been in place since 2005 and covers 75% of French industrial emissions.
  2. A carbon tax on fossil fuel consumption, starting at 7 euros per tonne of CO2 and now amounting to 45 euros per tonne, in place since 2014.

These increasingly stringent carbon pricing policies have taken place in a period of rising industrial energy costs generating concerns about their impact on the competitiveness of the manufacturing sector. At first glance, such concerns appear to be borne out. Recent trends show real output and total employment in the sector decreased by 5% and 26%, respectively, between 2001 and 2016.

However, a recent OECD report, shed another light on this issue. This study is the first to estimate the impact of energy prices and carbon taxes on the environmental and economic performance using data at the firm and industry level.

What does the OECD study tell us?

  1. At the firm level, a 10% increase in energy costs results in a 6% decline in energy use, a 9% decrease in carbon emissions, and a 2% decrease in the number of full-time employees within one year. However, these jobs are not lost, but are reallocated to other firms.
  2. At the industry level, there is no statistical link between energy prices and net job creation, indicating that jobs lost at affected firms are compensated by increases in employment in other firms operating in the same sector during the same year.

These effects vary both between industries and according to the size of the firm and their energy intensity: For example,

  • When facing the same increase in the energy cost, firms in the wearing apparel industry reduces their carbon emissions twice as much as firms producing non-metallic minerals.
  • Reallocation of workers in the food products industry is half the reallocation in the basic metals industry.
  • On average, large and energy intensive firms experience greater reduction in carbon emissions and greater job reallocation than smaller and energy efficient firms.

With this, the paper is able to measure the causal effect of the carbon tax on the aggregate manufacturing sector since its introduction in 2014. Figure 1 plots the carbon tax on the left axis (green line) together with the impacts of the carbon tax on the French manufacturing sector’s jobs (purple line) and carbon emissions (red line) on the right axis. In five years, the carbon tax decreased carbon emissions by 5%. The net effect on employment is much smaller in magnitude and even slightly positive at +0.8%.

Impact of the French carbon tax on aggregate jobs and CO2 emissions

Finally, the paper considers a scenario where the carbon tax is doubled from its current rate of 45 € per tonne of CO2. Figure 2 shows the simulated effect of the tax increase on job reallocations and carbon emissions for each manufacturing industry. These job reallocations are not net job losses, but the number of people forced to change jobs (within the same industry or between industries).

The impact of a doubling of the carbon tax on job reallocations and CO2 emissions

A simulated doubling of the carbon tax highlights significant heterogeneity across sectors. Several industries such as furniture, wood products, paper, and textiles experience large reductions in carbon emissions with little job reallocation. On the contrary, the motor vehicles and the plastic industries experience larger job reallocations and smaller declines in carbon emissions. Other industries such as metal products experience large job reallocation and emissions reduction because of their size.

Higher energy prices and carbon taxes are effective at reducing carbon emissions, but costs of job reallocation must be considered…

Although the carbon tax enables the French manufacturing sector to meet its carbon budget and does not affect total employment negatively, it however generates non-negligible job reallocations in several industries. Because these reallocation effects have redistributive implications and generate costs for workers who are forced to change jobs, these results call for complementary labour market policies that minimise those costs on affected workers and ease between-firms adjustments in employment. Moreover, since these transition costs are typically highly localised in regions specialised in polluting activities, they can also translate into potentially significant regional effects and thus political costs.


Dussaux, D. (2020), “The joint effects of energy prices and carbon taxes on environmental and economic performance: Evidence from the French manufacturing sector”, OECD Environment Working Papers, No. 154, OECD Publishing, Paris,

Taxe carbone : quel impact environnemental et économique dans le secteur manufacturier français ?

par Damien Dussaux, Economist, Direction de l’environnement, OCDE

En septembre 2019, le Parlement français a adopté la loi climat-énergie qui fixe l’objectif d’atteindre la neutralité carbone d’ici 2050, conformément à l’accord de Paris sur le climat de 2015. La neutralité carbone implique de réduire les émissions de gaz à effet de serre (GES) de 75 % d’ici 2050 par rapport aux niveaux de 1990 et de compenser les émissions résiduelles par la capture et le stockage du carbone présent dans l’atmosphère.

Afin d’atteindre cet objectif, le gouvernement français a élaboré une “Stratégie Nationale Bas-Carbone”, qui sert de feuille de route pour la transition vers de faibles émissions de carbone dans chaque secteur de l’économie. Par exemple, les émissions de GES de l’industrie – qui représentent près d’un cinquième des émissions en France, soit l’équivalent des émissions totales de GES de la Roumanie – devront être réduites d’un quart au cours des dix prochaines années selon le plan sectoriel proposé.

La France met actuellement en œuvre deux principaux mécanismes de tarification du carbone afin d’inciter les acteurs économiques à réduire leurs émissions. Le système européen d’échange de quotas d’émission (SEQE-UE), en place depuis 2005, couvre 75 % des émissions industrielles françaises. En 2014, la France a également introduit une taxe carbone sur la consommation de combustibles fossiles, qui a débuté à 7 euros par tonne de CO2 et s’élève désormais à 45 euros par tonne.

Ces politiques de tarification du carbone de plus en plus ambitieuses ont été mises en place dans un contexte de hausse des coûts énergétiques dans l’industrie et suscitent des inquiétudes quant à leurs impacts sur la compétitivité du secteur manufacturier français. À première vue, ces préoccupations semblent être justifiées par les tendances récentes, puisque la production et l’emploi total dans le secteur ont diminué respectivement de 5 % et 26 % entre 2001 et 2016.

Un récent rapport de l’OCDE, “Les effets conjugués des prix de l’énergie et de la taxe carbone sur la performance économique et environnementale des entreprises françaises du secteur manufacturier“, apporte un éclairage sur cette question. Cette étude est la première à estimer l’impact des prix de l’énergie et de la taxe carbone sur la performance environnementale et économique des entreprises françaises à partir de données à la fois au niveau des entreprises et au niveau des branches d’activité.

Le rapport combine des données sur la consommation d’énergie et les émissions de carbone au niveau des entreprises provenant de l’enquête annuelle sur la consommation d’énergie dans l’industrie (EACEI) de l’Insee avec les données sur les performances financières et économiques de la Direction Générale des Finances Publiques. L’ensemble des données couvre 8 000 entreprises françaises observées annuellement sur une période de seize ans (2001 à 2016) et qui sont représentatives de l’ensemble du secteur manufacturier.

Que nous apprend l’étude de l’OCDE ?

Le premier enseignement de l’étude est qu’au niveau des entreprises, une augmentation de 10 % des coûts énergétiques entraîne à court terme une baisse de 6 % de la consommation d’énergie, une diminution de 9 % des émissions de carbone et une diminution de 2 % du nombre d’employés à temps plein. Toutefois, ces emplois ne sont pas détruits car les salariés concernés sont embauchés dans d’autres entreprises. Au niveau de l’industrie, l’étude ne trouve aucun lien statistique entre les prix de l’énergie et la destruction nette d’emplois, ce qui indique que les emplois détruits dans les entreprises touchées sont compensés par des embauches dans d’autres entreprises de la même branche d’activité au cours de la même année.

Deuxièmement, ces effets varient d’une industrie à l’autre et en fonction de la taille et de l’intensité énergétique des entreprises. Par exemple, face à une même augmentation du coût de l’énergie, les entreprises de l’industrie de l’habillement réduisent leurs émissions de carbone deux fois plus que les entreprises produisant des minéraux non métalliques. Le redéploiement des travailleurs dans l’industrie agroalimentaire est deux fois moins important que dans l’industrie métallurgique. En moyenne, les grandes entreprises à forte intensité énergétique réduisent davantage leurs émissions de carbone et redéployent davantage de salariés que les petites entreprises efficaces en énergie.

Le rapport est ainsi en mesure de quantifier l’effet causal de la taxe carbone sur le secteur manufacturier depuis son introduction en 2014. La Graphique 1 montre la taxe carbone sur l’axe de gauche (ligne verte) ainsi que les impacts de la taxe carbone sur l’emploi total du secteur manufacturier français (ligne violette) et ses émissions de carbone (ligne rouge) sur l’axe de droite. En cinq ans, la taxe carbone a permis de réduire les émissions de carbone d’environ 5 %. L’effet net sur l’emploi est beaucoup plus faible et même légèrement positif à +0,8 %.

Enfin, le rapport envisage un scénario dans lequel le taux de la taxe sur le carbone serait doublé par rapport à son taux actuel de 45 € par tonne de CO2. Le Graphique 2 montre l’effet simulé de l’augmentation de la taxe sur les redéploiements de salariés et les émissions de carbone pour chaque branche d’activité. Ces redéploiements de salariés ne sont pas des pertes nettes d’emplois, mais le nombre de personnes contraintes de changer d’emploi (au sein d’une même industrie ou entre industries).

La simulation du doublement du taux de la taxe carbone met en évidence une grande hétérogénéité entre les branches d’activités. Plusieurs secteurs, tels que l’ameublement, les produits du bois, le papier et le textile, connaissent de fortes réductions de leurs émissions de carbone, avec un faible redéploiement de salariés. Au contraire, les secteurs de l’automobile et du plastique connaissent des redéploiements de salariés plus importants et des diminutions plus faibles de leurs émissions de carbone. D’autres industries, telle que celle des produits métalliques, combinent une forte réaffectation des emplois et une réduction considérable des émissions en raison de leur taille importante.

La hausse des prix de l’énergie et de la taxe carbone permet de réduire les émissions de carbone, mais les coûts liés aux redéploiements de salariés doivent être pris en compte…

Si la taxe carbone permet au secteur manufacturier français de respecter son budget carbone et n’affecte pas négativement l’emploi total, elle génère cependant des redéploiements de salariés non négligeables dans plusieurs branches d’activité. Parce que ces redéploiements ont des impacts redistributifs et génèrent des coûts pour les travailleurs qui sont contraints de changer d’emploi, ces résultats mettent en évidence la nécessité de mettre en place des politiques complémentaires sur le marché du travail qui minimisent les coûts pour les travailleurs concernés et facilitent les ajustements en termes d’emplois entre les entreprises. En outre, comme ces coûts de transition sont généralement fortement localisés dans des régions spécialisées dans les activités industrielles énergivores, ils peuvent également se traduire par des effets régionaux potentiellement importants et donc par un coût politique élevé.


Dussaux, D. (2020), “Les effets conjugués des prix de l’énergie et de la taxe carbone sur la performance économique et environnementale des entreprises françaises du secteur manufacturier”, OECD, No. 154, OECD Publishing, Paris,

Pollution Havens – just a delusion?

by Christina Timiliotis, Junior Trade Policy Analyst, OECD Trade & Agriculture Directorate, and Tomasz Kozluk, Head of the Green Growth Workstream, OECD Economics Department

Governments in the OECD and elsewhere must intensify efforts to mitigate pollution levels, if the international agreement of the latest COP 21 – pledging to keep global warming below 2 degrees – is to be more than just a loose promise. Against this background, policy makers need to enforce environmental regulations that oblige firms to account for the impact their actions have on the environment, and increase the price of using the environment as a factor of production. While there is broad support for environmental goals in the first place, support dwindles when compliance with such regulations implies higher production costs.

Efforts to put a cost on pollution have indeed often provoked resentment and resistance by producers and workers who fear to be put at a disadvantage vis-à-vis foreign competitors that are located in jurisdictions with laxer environmental policies. The conventional wisdom that tougher environmental regulations ultimately entail a loss in competitiveness and thus encourage industries to relocate production to a more favorable business environment, is commonly referred to as the “Pollution Haven Hypothesis” (PHH). If real, it can make environmental policy making politically difficult due to voters resistance and ineffective due to leakage. However, in spite of the PHH’s popularity, the evidence behind it is fragmented and to a large extent anecdotal.

The working paper “Do environmental policies affect global value chains? A new perspective on the pollution haven hypothesis” attempts to change this by looking at the Pollution Haven Hypothesis  through a new lens, using trade in value added data that more accurately represents today’s trade flows in the context of internationally fragmented value chains.

Scrutinising data across more and less pollution intensive industries in 23 OECD countries and six emerging economies since the 1990s, we find that countries with relatively stringent environmental laws do not suffer from lower exports as a result (see compare your country data viz) . There is however, a small effect on their relative competitiveness across different sectors in the economy. In countries with more stringent policies, exports of pollution and energy intensive sectors, such as steel-making or chemicals are lower than in the absence of stringent environmental policies (Figure 1). However, this is compensated by a corresponding increase in exports in “cleaner” industries like machinery or electronics. Moreover, both the positive and negative effects of environmental regulations on exports of different sectors have been small so far relative to the effects of other factors, such as market size, globalisation, national endowments or trade liberalisation.

Increase in domestic value added in exports 1995-2008, USD billions


Note: The figure shows exports from the three most stringent countries (Denmark, Germany, Switzerland) to BRIICS and vice versa, in billions USD. Pollution intensive sectors are defined according to methodology described in Kozluk and Timiliotis as ISIC rev. 3.1. 2325: Manufacture of coke, refined petroleum products and nuclear fuel; Manufacture of chemicals and chemical products; Manufacture of rubber and plastics products and 2000: Manufacture of wood and of products of wood and cork, except furniture; manufacture of articles of straw and plaiting materials. Less pollution intensive sectors are defined as 2933: Manufacture of machinery and equipment n.e.c.;  Manufacture of office, accounting and computing machinery; Manufacture of electrical machinery and apparatus n.e.c.; Manufacture of radio, television and communication equipment and apparatus; Manufacture of medical, precision and optical instruments, watches and clocks; 3637: Manufacture of furniture; manufacturing n.e.c.; Recycling.

In its preface to the General Theory, Keynes said that “the difficulty lies not so much in developing new ideas as in escaping old ones.” A myriad of ideas on how to credibly reduce the incentives to pollute in the long-term already exists. It remains to escape the archaic belief that ensuring environmental protection while maintaining a strong market position is infeasible. Governments must stand up to the environmental challenge and focus on the good design of environmental policies, accompanying framework policies and on the edge they can get from innovation – in order to secure both good environmental and economic outcomes.


Koźluk, T. and C. Timiliotis (2016), “Do environmental policies affect Global Value Chains? A new perspective on the pollution haven hypothesis”, OECD Economics Department Working Papers, No. 1282, OECD Publishing, Paris.

Further information:

Environment and trade: Do stricter environmental policies hurt export competitiveness?

Environmental Policies and Economic Performance, OECD Insights blog