FOCUS

Sustainable Finance: Current Needs, Measures and ImpactWilliam Oman and Romain Svartzman, Robin Döttling and Sehoon Kim, Caroline Flammer, Ethan Rouen and George Serafeim, Hao Liang and Jasper Teo, Christa Hainz, Johann Wackerbauer and Tanja Stitteneder

RESEARCH REPORT

Walking the Tightrope: Avoiding a Lockdown While Containing the VirusBalázs Égert, Yvan Guillemette, Fabrice Murtin and David Turner

REFORM MODEL

Discretionary Intervention Destabilizes the EU Emissions Trading System: Evidence and Recommendations for a Rule-Based Cap AdjustmentMichael Pahle and Ottmar Edenhofer

DICE DATA ANALYSIS

Tracking Government Responses to Covid-19: The CoronaNet Research ProjectCindy Cheng, Luca Messerschmidt, Svanhildur Thorvaldsdottir, Clara Albrecht, Christa Hainz, Tanja Stitteneder, Joan Barceló, Vanja Grujic, Allison Spencer Hart-nett, Robert Kubinec, Timothy Model and Caress Schenk

MACRO DATA INSIGHTS

Statistics Update €

32021

May Vol. 22

CESifo ForumISSN 1615-245X (print version)ISSN 2190-717X (electronic version)

A bi-monthly journal on European economic issuesPublisher and distributor: ifo Institute, Poschingerstr. 5, 81679 Munich, GermanyTelephone +49 89 9224-0, telefax +49 89 9224-98 53 69, email ifo@ifo.deAnnual subscription rate: €50.00Single subscription rate: €15.00Shipping not includedEditors: Yvonne Giesing, Christa Hainz, Chang Woon NamEditor of this issue: Christa HainzCopy editing: Clara Albrecht and Tanja StittenederIndexed in EconLitReproduction permitted only if source is stated and copy is sent to the ifo Institute.

www.cesifo.org

3/2021 FORUMSustainable Finance: Current Needs, Measures and Impact

Greening the economy is high on the political agenda. As early as 2015, the Paris Agreement set ambitious targets to combat climate change. Efforts have since become broader and are now aimed at promoting sustainable development in general, as set out in the UN 2030 Agenda for Sustainable Develop-ment. The financial sector is expected to play an important role in the upcoming transition towards a more sustainable economy. Therefore, policy makers in different parts of the

world are developing “sustainable finance” programs. The EU Commission has just released a renewed sustain-

able finance strategy. This issue of CESifo Forum explores why sustainable finance programs may be necessary and takes a closer look at investor and corporate behavior. The authors assess how important the objective of sustainability has been

so far and examine how it affects the decisions of companies and those of their peers. They also pro-

pose measures to render the corporate sector more sustainable.

FOCUS

Sustainable Finance: Current Needs, Measures and Impact

What Justifies Sustainable Finance Measures? Financial-Economic Interactions and Possible Implications for Policymakers 3William Oman and Romain Svartzman

ESG Investments and Investors’ Preferences 12Robin Döttling and Sehoon Kim

Short-Termism and Executive Compensation 17Caroline Flammer

Impact-Weighted Financial Accounts: A Paradigm Shift 20Ethan Rouen and George Serafeim

Peer Effects of Commitment to ESG Goals – How Stakeholders Affect One Another in the Ecosystem 26Hao Liang and Jasper Teo

Economic Policy Goals of the Sustainable Finance Approach: Challenges for SMEs 30Christa Hainz, Johann Wackerbauer and Tanja Stitteneder

RESEARCH REPORT

Walking the Tightrope: Avoiding a Lockdown While Containing the Virus 34Balázs Égert, Yvan Guillemette, Fabrice Murtin and David Turner

REFORM MODEL

Discretionary Intervention Destabilizes the EU Emissions Trading System: Evidence and Recommendations for a Rule-Based Cap Adjustment 41Michael Pahle and Ottmar Edenhofer

DICE DATA ANALYSIS

Tracking Government Responses to Covid-19: The CoronaNet Research Project 47Cindy Cheng, Luca Messerschmidt, Svanhildur Thorvaldsdottir, Clara Albrecht, Christa Hainz, Tanja Stitteneder, Joan Barceló, Vanja Grujic, Allison Spencer Hartnett, Robert Kubinec, Timothy Model, Caress Schenk

Statistics Update 51

3CESifo Forum 3 / 2021 May Volume 22

FOCUS

Sustainable Finance: Current Needs, Measures and ImpactGreening the economy is high on the political agenda. As early as 2015, the Paris Agreement set ambitious targets to combat climate change. Efforts have since become broader and are now aimed at promoting sustainable development in general, as set out in the UN 2030 Agenda for Sustainable Development. The financial sector is ex-pected to play an important role in the upcoming transition towards a more sustain-able economy. Therefore, policy makers in different parts of the world are develop-ing “sustainable finance” programs. The EU Commission has just released a renewed sustainable finance strategy. This issue of CESifo Forum explores why sustainable finance programs may be necessary and takes a closer look at investor and corpo-rate behavior. The authors assess how important the objective of sustainability has been so far and examine how it affects the decisions of companies and those of their peers. They also propose measures to render the corporate sector more sustainable.

William Oman and Romain Svartzman

What Justifies Sustainable Finance Measures? Financial-Economic Interactions and Possible Implications for Policymakers

Climate change looms increasingly large on the policy agenda. The evidence shows that climate change is accelerating and its effects are becoming ever more severe (Slater et al. 2021). Yet, according to the United Nations, countries’ updated climate pledges would re-sult in emissions that are only 0.5 percent below 2010 levels by 2030—far below the 45 percent reduction that the Intergovernmental Panel on Climate Change (IPCC) views as necessary to limit global warming to 1.5°C (IPCC 2018). If policies remain unchanged, global temperatures could rise by a further 2–5°C by 2100 (levels not seen in millions of years), raising the risk of catastrophic outcomes (IMF 2020a). Patricia Espinosa, the executive secretary of the UN Framework Conven-tion on Climate Change (UNFCCC), put it bluntly: “We are collectively walking into a minefield blindfolded. The next step could be disaster.”

At the same time, policymakers seem increasingly aware of the need to act with urgency and determi-nation, although in practice the measures taken are still far from sufficient. Solving these challenges will require dramatic changes in production, consumption

and distribution patterns as well as lifestyles (IPCC 2018), and in some cases very large investments.

It is against this backdrop that work on climate-related financial risks and sustainable finance measures has been in-tensifying. Sustainable finance measures can be defined as pol-icies that either seek to or can directly or indirectly help align the financial system with environ-mental goals. They include both financial-sector policies, notably financial supervision and regula-tion, and monetary policies, as both affect financial markets, financial actors and financial assets, and ultimately the allo-cation of capital. Following Mark Carney’s landmark speech on the “tragedy of the horizon” (Carney 2015), sustainable finance moved to center stage in the policymak-ing community, in particular with the creation in 2017 of the Net-

1 The views expressed in this paper are those of the authors and do not necessarily represent the views of the Banque de France, the IMF, its Executive Board, or IMF management. The authors thank Jean Boissinot, Clément Bourgey and Dora Iakova for their comments and suggestions.

is Economist at the International Monetary Fund. He focuses on macrofinancial issues and climate change within the European Department.

William Oman

Romain Svartzman

is Economist at the Banque de France. He focuses on the as-sessment of nature-related finan-cial risks and on sustainable fi-nance within the Climate Change Centre.

CONTENT

4 CESifo Forum 3 / 2021 May Volume 22

FOCUS

work of Central Banks and Supervisors for Greening the Financial System (NGFS). Reflecting the growing consensus in the central banking community on the significance of climate change for financial stability, the NGFS comprises 87 members (as of the time of writing), including the world’s main central banks.

While such developments are welcome, the ra-tionale for promoting sustainable finance remains under debate. In particular, according to standard economic theory, in which the financial sector’s role is to smooth income over time, pool savings and chan-nel them to productive investment, process and share information, and diversify risk (Merton 1995), focusing on the financial sector would be inefficient and inef-fective as long as measures are not taken in the real economy, most importantly a significant increase in carbon prices (IMF 2019). However, other theoretical and practical approaches, including some put forward by central banks and financial supervisors, suggest that the financial sector may have an important role to play in decarbonizing the economy.

This article provides an overview of the main jus-tifications for sustainable finance measures that have been proposed to date, and their related theoretical and practical challenges. It does so by taking stock of the literature and recent policy developments, and by focusing on the role of central banks and financial regulators and supervisors. The next section summa-rizes the theoretical and practical justifications that have been given for sustainable finance measures. The following section considers barriers to the alignment of financial flows with sustainability goals and pro-posals to overcome these limitations. The concluding section discusses two ongoing debates and challenges related to the development of sustainable finance and the role of central banks.

JUSTIFICATIONS FOR SUSTAINABLE FINANCE MEASURES: FROM ECONOMIC THEORY TO PRACTICAL CONSIDERATIONS

Justifications for sustainable finance measures can be divided into two broad categories: theoretical and practical. Theoretical justifications focus on external-ities that hinder the low-carbon transition. Practical justifications highlight two potential threats to policy objectives posed by climate change: practical limita-tions to the concept of externalities, and the need for central banks and financial supervisors to manage the systemic risks generated by climate change and, potentially, other ecological crises. Below, we discuss each type of justification in turn, while acknowledging that they can be complementary rather than exclusive.

Sustainable Finance: Auxiliary, Second Best or Essential Actor of the Low-Carbon Transition?

The standard approach to environmental problems in economics has been to define the latter as negative

externalities, that is, activities that—if not corrected by policies—have a direct negative impact on others’ production and consumption possibilities, including those of future generations. Taking this view, the emission of greenhouse gases (GHGs) imposes the externality of climate change (Stern 2015). As such, the role of the policymaker consists in ensuring that markets reinternalize the externality. To do so, the first-best solution consists in a carbon price, usually in the form of a Pigouvian tax on emissions that is equal to the external damage (Pigato et al. 2019). Other in-struments exist, such as emission trading schemes (ETS), also known as cap-and-trade systems. Unlike a tax, which sets the price and allows the CO2 level to vary, a cap-and-trade system sets the overall CO2 level and allows the price to be determined in a decentral-ized way. In a first-best world, the financial sector does not play a role in decarbonizing the economy. The task of policy is to use carbon pricing to correct the GHG externality, which is assumed to be the only existing market failure, as laid out in Stern (2007). In such a world, sustainable finance can contribute to increasing the effectiveness of climate policies and help better manage environmental risks; aiming at greening the financial system without taking meas-ures to green the ‘real’ economy can, though, lead to greenwashing (NGFS 2020a).

However, the view that carbon pricing alone will suffice suffers from serious limitations. The GHG ex-ternality is unique because climate change is global in its scope and impact, involves a high level of un-certainty, and is long-term and governed by a stock-flow process that makes it difficult to react quickly if mistakes are made, not least because its effects are potentially huge and irreversible. As a result, The Stern Review famously called climate change the “greatest market failure the world has ever seen” (Stern 2007). Empirical evidence shows that many economic sec-tors display a low elasticity of emissions to carbon prices (Rafaty et al. 2020), particularly in sectors with structural challenges such as urban systems, indus-trial supply chains and production networks (Hep-burn et al. 2020). Stern and Stiglitz (2021) argue that climate change “involves radical change in all of the core systems of the economy (e.g. energy, land, cit-ies, transportation).” This suggests that, while nec-essary, carbon pricing alone is insufficient to induce the necessary structural change to decarbonize the global economy in the required timeframe. Instead, achieving the objective of the Paris Agreement of lim-iting global warming to well below 2 degrees Celsius requires countries to implement policy packages that include policies beyond carbon pricing to address the multiple market failures that can undermine the low carbon transition (High-Level Commission on Carbon Prices 2017; Krogstrup and Oman 2019; Bhattacharya et al. forthcoming).

In this context, a second complementary theoret-ical justification for sustainable finance measures can

CONTENT

5CESifo Forum 3 / 2021 May Volume 22

FOCUS

be envisioned, whereby measures aimed at promot-ing sustainable finance could be considered as sec-ond-best policies. According to Lipsey and Lancaster (1956), second-best policies are justified when there are multiple market failures that cannot be corrected independently, such that correcting one externality can reduce welfare. For instance, those on the losing side of climate change may not have the resources to make the productive investments needed, and/or imperfections in capital markets may mean that the latter will not provide the necessary low-carbon finance, meaning that sustainable financial measures become necessary (Stern and Stiglitz 2021; Bhattacha-rya et al. forthcoming).

However, the limitations to the pricing of exter-nalities may be due to the fact that centuries of in-vestment in fossil fuels have led to considerable in-stitutional and technological inertia, cementing the structure of the global economy (Bhattacharya et al. forthcoming). Going further, it has been argued that market economies regularly generate new ex-ternalities, such that regulatory systems are overrun by externalities (Kapp 1950). According to this view, climate change should not be approached as a mar-ket failure (no matter how massive) but rather as a systemic challenge that calls for an unprecedented level of coordination among, and commitment from, multiple actors (private and public), involving multi-ple instruments (pricing, sectoral regulations, and so on), and with multiple consequences (on inequality, for instance).

These considerations point towards a third theo-retical underpinning for sustainable finance, in which, more than being second best, sustainable finance could in fact have a central role to play in addressing climate change. Indeed, finance is critical for funding the new kinds of innovation and investments that are needed for deep decarbonization (Fay et al. 2015). Some estimates have found that low-carbon infra-structure investment gaps in developing countries could reach USD 15-30 trillion by 2040, and capital is not being allocated in a way that is consistent with the goals of the Paris Agreement (Bayat-Renoux et al. 2020, IPCC 2018). Significant public investment is key, but large-scale private investment is also needed to achieve the required structural transformation away from carbon-intensive economic systems towards a net zero economy (Villeroy de Galhau 2015). It is, in fact, notable that the role of the financial system has been explicitly acknowledged in climate negotiations, with the Paris Agreement calling for “making finance flows consistent with a pathway towards low green-house gas emissions and climate-resilient develop-ment” (UNFCCC 2016).

This need for sustainable finance is all the more urgent since, according to many observers, in its cur-rent form and without high carbon prices, the global financial system hinders rather than enables the deep decarbonization needed (IPCC 2018; Bayat-Renoux

et al. 2020).2 Stern and Stiglitz (2021) outline four critical market failures and so-called government failures underpinning sustainable finance measures: (i) capital market imperfections, in particular highly imperfect climate risk markets, which lead to credit rationing for low-carbon investments; (ii) socialization of losses, which leads to collective moral hazard and excessive risk taking (such as fossil fuel investments that greatly risk becoming stranded assets); (iii) com-mitment problems, notably the lack of credibility of a ‘no bail-out’ rule for large fossil-fuel firms; (iv) sys-tematic short-termism in managerial incentives due to imperfect and asymmetric information, which leads to managerial decisions that entail short-term benefits to managers and excessive climate risks, thereby leading to the famed “tragedy of the horizon” (Carney 2015). Some of these justifications are particularly relevant for developing countries, which have considerable long-term financing needs but limited access to cli-mate finance (leading to higher hurdle rates), espe-cially in the context of the current crisis (Bayat-Renoux et al. 2020; Svartzman and Althouse 2020).

What Role for Central Banks and Financial Supervisors? The Risk-Based Approach

This three-pronged conceptual framework helps us understand why central banks and financial supervi-sors have started to pay attention to climate change by collaborating through the NGFS, and how they rap-idly became stakeholders of the low-carbon transition within the financial sector. The main contribution of the NGFS if to consider that, insofar as climate change poses a risk to financial stability, it “falls squarely within the mandates of central banks and supervi-sors to ensure the financial system is resilient to these risks” (NGFS 2019).

NGFS members focus on two types of climate-re-lated financial risks: physical and transition risks (Ban-que de France, ACPR, and DG Trésor 2017; NGFS 2019; Grippa et al. 2019). Physical risks correspond to finan-cial losses resulting from more frequent and severe weather and climate extremes (e.g., storms, wildfires) and long-term changes in climate patterns (e.g., rising sea levels). Transition risks correspond to financial losses resulting from a rapid or disorderly low-carbon transition. The latter could be triggered or accelerated by policy changes, technological disruptions, or be-havioral changes. Such shifts could generate ‘stranded assets,’ especially in carbon-intensive sectors tied to fossil fuels (McGlade and Elkins 2015; Mercure et al. 2018). These stranded assets could in turn expe-rience a sudden repricing, potentially causing a “cli-mate Minsky moment” (Carney 2015). Physical and transition risks could interact with one another and materialize in many different manners while generat-

2 According to estimates by the World Resources Institute, the world’s largest 33 banks allocated USD 654 billion to fossil fuel fi-nancing in 2019 (Avery 2019).

CONTENT

6 CESifo Forum 3 / 2021 May Volume 22

FOCUS

ing feedback loops that can affect all economic agents (Figure 1). The implications of climate change for price stability are also being considered (e.g., Parker 2018), with some central banks exploring how their primary mandate of price stability could be impacted by cli-mate change (Cœuré 2018; NGFS 2020b; Lagarde 2020; Schnabel 2021; Villeroy de Galhau 2021).

To identify and manage such climate-related financial risks, the emerging consensus among central banks is that it is necessary to rely on for-ward-looking, scenario-based analyses (Aufauvre and Bourgey 2019; Campiglio et al. 2018; NGFS 2020c). Unlike backward-looking risk management approaches, scenario analyses seek to examine plau-sible hypotheses for the future without assigning probabilities. Building on macrofinancial stress tests used to assess the resilience of banks in adverse sce-narios (Borio et al. 2014), some central banks and the IMF are now developing adverse climate scenarios that will lead to “climate stress tests” (ACPR/Ban-que de France 2020; Vermeulen et al. 2019; Adrian et al. 2020)—that is, to estimating the resilience of the financial system to specific climate-related phys-ical and transition developments (see Battiston and Monasterolo 2017).

These developments have triggered intense de-bates related to the theoretical considerations dis-cussed above, revolving around whether and how central banks should play a role in combating climate change. Indeed, many have argued that central banks may end up overstepping their mandate by focusing on policy objectives that are within the purview of elected officials (see, e.g., Pisani-Ferry 2021). How-ever, central banks’ position escapes this critique. The question being asked by NGFS members is not whether they should replace elected policymakers, most notably the fiscal authorities, in implement-ing climate policy (as noted by Weidmann 2021), but

rather how they should integrate climate-related fi-nancial risks into an appropriate risk management strategy that enables central banks to preserve both price and financial stability in the age of climate change (NGFS 2020b).

It thus appears that the NGFS’ position is mostly grounded in the third approach proposed above, while acknowledging its interactions with the first two approaches (as further discussed below), which considers that action from all actors will be needed to tackle climate change. Central banks and financial supervisors are part of the set of relevant actors, therefore they can and should act within the remit of their mandates. This position is informed by the view that assessing climate-related risks will itself contribute to decarbonizing the economy. As noted by Carney (2015), the approach of central bankers to both financial and price stability in the age of climate change has been similar so far: it is necessary to better measure climate-related risks so as to manage them. In other words, the impacts of climate change should be managed according to the “old adage […] that which is measured can be managed.”

CHALLENGES FOR THE ALIGNMENT OF FINANCIAL FLOWS WITH SOUND CLIMATE-RELATED RISK MANAGEMENT

While sustainable finance measures have focused on climate-related risks, some observers have warned that these risks may in fact be impossible to measure (even if one were able to fully bridge existing data and information gaps) and/or to manage from a pure risk-based approach. Below, we review these arguments and discuss some possible implications for central banks’ theoretical framework with respect to climate change.

Can Climate-Related Financial Risks be Measured and Managed?

The uncertainty around climate change poses a chal-lenge to the measurement of climate-related financial risks. With regard to physical risks, tipping points are very likely to exist within Earth ecosystems, but re-main difficult to estimate, and exceeding them could generate multiple cascade reactions (Lenton et al. 2019; Steffen et al. 2018) that make them particularly difficult to translate into financial metrics over un-certain time horizons. For example, while it is com-monly agreed that climate change could generate mass migrations and conflicts (Abel et al. 2019), the probability of occurrence of such events and their translation into social, economic and then financial metrics are inherently difficult to measure with any degree of confidence.

Uncertainty is also pervasive when it comes to assessing transition risks. For instance, it remains highly uncertain which technologies will prevail in

Source: Bolton et al. (2020a).

Channels and Spillovers for the Materialization of Physical and Transition Risks

© ifo Institute

Source of risk Transmission channels

Impacts on sovereigns

Impacts on households

Impacts on sectors’/firms’:

– sales– operational costs– CAPEX– asset and equity valuation– etc.

Financial risks and contagion

Credit risks

Market risks

Liquidity risks

Insurance risks

Operational risks

Feedback loops

Transition risks: policy, technology,

social norms and preferences

Physical risks: extreme weather events,

long-term changes in climate patterns

Figure 1

CONTENT

7CESifo Forum 3 / 2021 May Volume 22

FOCUS

a low carbon world (Svartzman et al. 2020). This makes any assessment of which countries, sectors and firms will win or lose from the low-carbon tran-sition particularly difficult. The future trajectory of carbon prices is also highly uncertain, as is the fu-ture value of avoided emissions (Aglietta et al. 2015). Such difficulties can be compounded if one takes into account the increasingly acknowledged fact that a low-carbon transition could reshuffle trade flows (e.g., because of the new materials needed to power renewable technologies) and trigger new geopolitical alliances and tensions (Tänzler and Gordon 2020). In short, risk-management approaches may be particu-larly challenged when it comes to measuring risks that will impact all agents and interact with multiple other dynamic patterns.

These considerations have led some to argue that policymakers should instead embrace the concept of deep (Boissinot and Heller 2020) or radical (Chenet et al. 2021) uncertainty. A central idea in macroeconom-ics is that uncertainty about the future is ubiquitous (Knight 1921; Keynes 1936). With regard to climate change, this uncertainty is particularly deep, and as Stern and Stiglitz (2021) note, “there may be states of nature that we have never experienced or find hard to imagine or describe.”3 Ultimately, this has led to the argument that the financial system and its robust-ness are “vulnerable not only to calculable risks but to ‘risks without a known distribution: so-called ‘Knigh-tian’ or deep uncertainty or ‘unknown unknowns’” (Zenghelis and Stern 2016).4

Even if climate-related financial risks could be measured, it is not clear that it would be possible to manage them. The idea of relying on increased transparency on climate risks assumes that manda-tory disclosure of climate risks will elicit an “efficient market reaction to climate change risks” (Carney 2015, see also FSB 2015).5 According to this logic, climate risk disclosure, if introduced early enough and pertinent and internationally consistent, should induce transparency and market discipline, which should raise the cost of capital for climate-risky assets at the time the capital is deployed, in turn fostering financial stability by helping avoid a “cli-mate Minsky moment.” It has been argued that this

3 Some observers have argued that assessing climate futures re-quires an assessment of the socio-political and socio-economic fu-tures entailed by climate change, which requires engaging with ‘deep,’ ‘fundamental,’ or ‘radical’ uncertainty (Kandlikar et al. 2005).4 A risk minimization approach (Krogstrup and Oman 2019) stresses the asymmetry of the costs of policy mistakes. The cost of acting too slowly to combat climate change is much greater than the cost of mitigating climate change ’too fast‘ – policymakers can reverse miti-gation actions, but they cannot reverse climate overshoot.5 Schnabel (2020) thus argues that the empirical evidence suggests a mispricing of climate risks “as a result of informational market fail-ures that stem primarily from the absence of a clear, consistent and transparent globally agreed taxonomy accompanied by disclosure requirements.” There is evidence that equity valuations do not re-flect the projected incidence of climate risks (IMF 2020b). Bolton and Kacperczyk (forthcoming) document that stocks of firms with higher emissions earn higher returns. ESRB (2020) find evidence that finan-cial market pricing of climate risks seems “heterogeneous at best, and absent at worst.”

approach implicitly assumes that financial markets are informationally efficient, a hypothesis that has been challenged in the literature.6

A New Dilemma for Central Banks, and a Potential Way Forward Through the Concept of Double Materiality?

The existence of widespread and significant barriers to the measurement and management of climate-re-lated financial risks poses a dilemma for central banks (Bolton et al. 2020a). On the one hand, for central banks to develop scenarios while waiting for other institutions or economic agents to take action (e.g., by imposing a carbon price) might expose them to the risk of not being able to deliver on their mandates of price and financial stability. On the other hand, central banks cannot substitute for other actors’ in-sufficient actions, as the latter pertain to areas that are not within the central bank’s remit (e.g., fiscal, industrial, urban planning, etc.).

To address this dilemma a new paradigmatic approach may be emerging through the concept of ‘double materiality,’ which is already supported by some policymakers, including financial supervisors (see, e.g., European Commission 2019 and ESMA 2020). Double materiality (see Figure 2) suggests that a comprehensive approach to climate-related finan-cial risks calls for assessing two related phenomena: the fact that climate change can affect financial insti-tutions (as captured by the risk-based approach), and the fact that financial institutions impact the climate system and therefore contribute to the risks they aim to measure. Some policymakers have hinted that the concept may apply to central banks, whose actions should not “reinforce market failures that threaten to slow down the decarbonization objectives of the global community” (Schnabel 2020). This idea ap-pears to have support in the literature, in particular the notion that the more climate action is delayed, the more skewed the distribution of climate sensitiv-

6 See Christophers (2017) for a more detailed discussion. See also Stern and Stiglitz (2021).

Double Materiality

Source: Authors’ illustration, based on European Commission (2019). © ifo Institute

Contribution to physical and transition risks

Vulnerability to physical and transition risks

BANK

Figure 2

CONTENT

8 CESifo Forum 3 / 2021 May Volume 22

FOCUS

ity values, and the higher the probability of extreme values (Weitzman 2012; Aglietta and Espagne 2016).7

In fact, the concept of double materiality seems also to be implicitly present in the purpose of the NGFS, as the latter seeks to “contribute to the de-velopment of environmental and climate risk man-agement in the financial sector and to mobilize the mainstream finance to support the transition toward a sustainable economy.”8 While the first part of the NGFS mission statement corresponds to identify-ing how climate change could impact the financial system, the second part corresponds to reducing the contribution of the financial system to climate change.

The implementation of the concept of dou-ble materiality could have diverse implications for prudential and monetary policy (e.g., Schoenmaker 2021, Van ‘t Klooster and Van Tilburg 2020). Monetary policy could be decarbonized in a gradual and tar-geted manner (Villeroy de Galhau 2021), for instance, by ensuring that the collateral pledged by the central bank’s counterparties is aligned (or consistent) with the objectives of the Paris Agreement based on cred-ible third parties’ opinions (see Oustry et al. 2020). A related proposal is to use taxonomies of activities to determine whether to exclude certain bonds, based on clear and transparent rules that are used to fi-nance projects that conflict with decarbonization objectives (Schnabel 2020). In this approach, cen-tral banks and financial supervisors would seek to steer markets in a low-carbon direction, not because they are acting beyond their mandate, but because the nature of climate-related financial risks requires them to broaden their approach to risks. This means that the double materiality perspective may lead to the need to consider whether reducing the impact of the financial system on the climate system could be part of a robust risk-management strategy.

SOME DEBATES AHEAD

While the above arguments may provide central banks with a theoretical and operational framework of action with regard to climate change, they would leave several questions unaddressed. We briefly discuss two of these below. First, given that cent- ral banks’ actions would remain inefficient in a world that does not act on climate change, what role, if any, is there for coordinating their actions with other actors (notably fiscal authorities) and what are the potential implications for central banks? Second, how do they account for environmental risks beyond climate change? Our aim is merely to present these debates, as they are likely to shape future discussions.

7 A related idea is that climate change is uninsurable (see Chichilni-sky and Heal 1993).8 See www.ngfs.net.

Policy Coordination Challenges and Potential Implications for Central Banks

In order to overcome the above dilemma (i.e., the need for central banks to act and the impossibility of solving the climate challenge on their own), one existing proposal consists in emphasizing policy coor-dination. As noted by the OECD (2015), it is necessary to make all policies consistent with climate objectives, and the role of central banks and sustainable finance policies should be assessed in this light. Stern and Stiglitz (2021) argue that climate change mitigation requires radical change in all of the core systems of the economy, which in turn requires “complex coor-dination of a kind that goes beyond standard pricing, especially in the presence of multiple market failures.”

As an example of this approach, Aglietta and Valla (2021) call for a transformation of the growth regime in which fiscal, financial and monetary policies are complementary and cooperative. They, among others (e.g., Gabor 2021), propose new thinking and a new organization of monetary policy, in which the latter is integrated with macroprudential policy and coor-dinated with fiscal policy. More broadly, Bolton et al. (2020a) identify four areas (fiscal policy, responsible investment, international coordination, and account-ing norms) that do not fall within the remit of central banks, but with which central banks may need to in-teract in order to manage climate-related financial risks.

Such avenues nevertheless pose at least three sig-nificant challenges. First, they require central banks’ to know what other actors will do, to enable them to adjust their own actions. For instance, knowing how climate policy will affect the economic outlook, and its implications for monetary policy (NGFS 2020b). Another example relates to the path and speed of the transition induced by climate policy, which may also have financial stability implications (NGFS 2020c).

Second, some of the measures mentioned above may be considered to impinge on the principle of mar-ket neutrality, according to which monetary policy should be asset neutral. Weidmann (2020) argues that it is not the task of the central bank to “penalize or promote certain industries.” Other senior central bankers have nevertheless questioned whether the principle itself should be reassessed in the light of climate change. For instance, Schnabel (2020) argues that “market neutrality may not be the appropriate benchmark for a central bank when the market by it-self is not achieving efficient outcomes” and Lagarde wonders “whether market neutrality should be the ac-tual principle that drives our monetary-policy portfo-lio management.”9 Villeroy de Galhau (2021) suggests that “market neutrality – which guides the execution

9 See Arnold (2020). Further, Schnabel (2021) notes that large firms in emission-intensive sectors are more likely to enter the bond mar-ket as they have a high level of fixed assets that can serve as collater-al.

CONTENT

9CESifo Forum 3 / 2021 May Volume 22

FOCUS

of our market operations – should not put a brake on carbon neutrality.”

Third, certain forms of policy coordination (e.g., fiscal-monetary) to achieve climate-related goals raise questions about central bank independence. Cochrane (2020) argues that by engaging in climate policy, central banks will lose their independence and ability to fulfill their main missions of controlling in-flation and stemming financial crises. By contrast, Honohan (2019) argues that failing to green central bank interventions will over the longer term pose a threat to central bank independence, while Bernanke (2003) has argued that, under some circumstances, “greater cooperation for a time between the central bank and the fiscal authorities is in no way inconsist-ent with the independence of the central bank.” More broadly, such questions relate to that of determining whether central banks should act within their primary or secondary mandate (in the case of the ECB), and the extent to which climate risks and mitigation goals fit into current central bank mandates (see Cœuré 2018, Dikau and Volz 2019; Fischer 2019; Van Tilburg and Simic 2021).

Beyond Climate-Related Financial Risks, A New Era of Ecological Risks?

While sustainable finance has mainly focused on cli-mate change, it has been argued that the latter is only the “tip of the iceberg” (Steffen et al, 2011). Other biogeochemical cycles that are as essential for life on Earth as a stable climate are also increasingly af-fected by human activity (e.g., biodiversity loss and soil erosion – IPBES 2019; IPCC 2019). Rockstrom et al. (2009) and Steffen et al. (2015) identify nine critical Earth systems with environmental boundaries that have already been crossed or could be crossed in the near future. Crossing these boundaries could lead to catastrophic outcomes for ecosystems and human systems (Lenton et al. 2019), let alone for economic and financial systems. For instance, a growing body of literature (Johnson et al. 2020) and international organizations (IPBES 2020) indicate that because of ongoing biodiversity loss, we may have entered an era of biodiversity-related financial risks (including pan-demics triggered by zoonotic diseases), which could make the occurrence of systemic and irreversible fi-nancial risks more likely (Bolton et al. 2020b; Dasg-puta 2021). Kedward et al. (2020) suggest capturing these large, interconnected risks under the notion of “nature-related financial risks.”

Some central banks have begun acknowledging some of these risks (Schnabel 2021; Schellekens and Van Toor 2019), in particular with regard to biodi-versity (DNB 2020). However, measuring biodiversi-ty-related financial risks is more complex than cli-mate-related risks, as the former relies on multiple local indicators related to the functioning of diverse ecosystems (Kedward et al. 2020). Central banks

may also be more constrained in their ability to ad-dress some of these issues than they are with climate change. For instance, a rapidly growing literature sug-gests that protecting ecosystems calls for large-scale transformations of socio-ecological systems, requir-ing the need for new tools to measure welfare (e.g., by moving beyond GDP and considering the limits to its growth) (Dasgupta 2021; IMF 2021). Although the literature on such risks is in its infancy, it is conceiv-able that the above considerations could ultimately reinforce the need for new institutional arrangements among policy areas. While central banks are likely to be exposed to risks from the systematic degradation of ecosystems and should therefore pay increasing attention to them, their margin of action will remain limited if they act on their own.

REFERENCES Abel, G. J., M. Brottrager, J. Crespo Cuaresma, and R. Muttarak (2019), “Climate, Conflict and Forced Migration,” Global Environmental Change 54, 239–49.

ACPR/Banque de France (2020), “Scénarios et hypothèses principales de l’exercice pilote climatique,” Autorité de Con-trôle Prudentiel et de Résolution, Direction d’études et d’analyse des risques, https://acpr.banque-france.fr/scenarios-et-hypotheses-principales-de-lexercice-pilote-climatique.

Adrian, T., J. Morsink, and L. Schumacher (2020), “Stress Testing at the IMF,” International Monetary Fund, Monetary and Capital Markets De-partment (Series) No. 20/04.

Aglietta, M., E. Espagne, and B. Perrissin Fabert (2015), “A Proposal to Finance Low-carbon Investment in Europe,” La Note D’Analyse No. 4, France Stratégie, Paris.

Aglietta, M. and E. Espagne (2016), “Climate and Finance Systemic Risks, More Than an Analogy? The Climate Fragility Hypothesis,” CEPII Working Paper no. 2016-10.

Aglietta, M. and N. Valla (2021), Le Futur de la monnaie, Odile Jacob, Paris.

Arnold, M. (2020), “ECB to Consider Using Climate Risk to Steer Bond Purchases, Says Lagarde,” Financial Times, October 14, https://www.ft.com/content/f5f34021-795f-47a2-aade-72eb5f455e09.

Aufauvre, N. and C. Bourgey (2019), “The Role of Central Banks and Su-pervisory Authorities in Stimulating Consideration for Long-term Issues: The Case of Climate Change,” Annales Des Mines - Réalités Industri-elles, 4, 60–63.

Avery, H. (2019), “Sustainable Finance’s Biggest Problems, By the People Who Know Best,” Euromoney, December 3, https://www.euromoney.com/article/b1j97rjr74vd00/sustainable-finances-biggest-problems-by-the-people-who-know-best.

Banque de France, ACPR, and DG Trésor (2017), “Assessing Climate- Change Related Risks in the Banking Sector.”

Battiston, S., A. Mandel, I. Monasterolo, F. Schütze, and G. Visentin (2017), “A Climate Stress-Test of the Financial System,” Nature Climate Change 7(4), 283–88.

Bayat-Renoux, F., H. de Coninck, Y. Glemarec, Y., J.-C. Hourcade, K. Ramakrishna, and A. Revi (2020), “Tipping or Turning Point: Scaling Up Climate Finance in the Era of COVID-19,” Green Climate Fund Working Paper 3.

Bernanke, B. (2003), “Some Thoughts on Monetary Policy in Japan,” Tokyo, May.

Bhattacharya, A., M. Ivanyna, W. Oman, and N. Stern (forthcoming), “Cli-mate Action to Unlock the Inclusive Growth Story of the 21st Century,” in How to Achieve Inclusive Growth (Cerra, V., B. Eichengreen, A. El-Ga-nainy, and M. Schindler, eds.), Oxford University Press, New York.

Boissinot, J. and T. C. Heller (2020), “Scénarios Climatiques en Finance,” Revue d’Economie Financière 138, 55–68.

Bolton, P., M. Després, L. A. Pereira da Silva, F. Samama, and R. Svartz-man (2020a), The Green Swan, Central Banking and Financial Stability in the Age of Climate Change, Bank for International Settlements and Banque de France.

CONTENT

10 CESifo Forum 3 / 2021 May Volume 22

FOCUS

Bolton, P., M. Després, L. A. Pereira da Silva, F. Samama, and R. Svartz-man (2020b), “Penser la Stabilité Financière à L’ère des Risques Écologiques Globaux - Vers de Nouveaux Arbitrages entre Efficience et Résilience des Systèmes Complexes”, Revue d’économie financière 138, 43–54.

Bolton, P. and M. T. Kacperczyk (forthcoming), “Do Investors Care about Carbon Risk?”, Journal of Financial Economics.

Borio, C., M. Drehmann, and K. Tsatsaronis (2014), “Stress-Testing Macro Stress Testing: Does It Live up to Expectations?”, Journal of Financial Stability 12(1), 3–15.

Campiglio, E., Y. Dafermos, P. Monnin, J. Ryan-Collins, G. Schotten, and M. Tanaka (2018), “Climate Change Challenges for Central Banks and Financial Regulators,” Nature Climate Change 8(6), 462–68.

Carney, M. (2015), “Breaking the Tragedy of the Horizon – Climate Change and Financial Stability,” Speech at Lloyd’s of London, Septem-ber 29.

Chenet, H., J. Ryan-Collins, and F. van Lerven (2021), “Finance, Climate- change and Radical Uncertainty: Towards a Precautionary Approach to Financial Policy,” Ecological Economics 183.

Chichilnisky G. and G. M. Heal (1993), “Global Environmental Risks,” Journal of Economic Perspectives 7(4), 65–86.

Christophers, B. (2017), “Climate Change and Financial Instability: Risk Disclosure and the Problematics of Neoliberal Governance,” Annals of the American Association of Geographers 107(5), 1108–1127.

Cochrane, J. H. (2020), “Central Banks and Climate: A Case Of Mission Creep,” Hoover Institution.

Cœuré, B. (2018), “Monetary Policy and Climate Change,” Speech given at a conference on “Scaling up Green Finance: The Role of Central Banks,” organized by the Network for Greening the Financial System, the Deutsche Bundesbank and the Council on Economic Policies, Berlin, November 8.

Dasgupta, P. (2021), The Economics of Biodiversity: The Dasgupta Review, London, HM Treasury.

Dikau, S., and U. Volz (2019), “Central Banking, Climate Change and Green Finance,” in Sachs, J., W. T. Woo, N. Yoshino and F. Taghizadeh- Hezary (eds.), Handbook of Green Finance: Energy Security and Sustain­able Development, Asian Development Bank Institute.

DNB (2020), “Indebted to Nature: Exploring Biodiversity Risks for the Dutch Financial Sector,” De Nederlandsche Bank.

ESMA (2020), “ESMA Supports IFRS Foundation’s Efforts on International Standardisation in Sustainability Reporting,” December 16, https://www.esma.europa.eu/press-news/esma-news/esma-supports-if-rs-foundation%E2%80%99s-efforts-international-standardisation-in.

ESRB (2020), “Positively Green: Measuring Climate Change Risks to Fi-nancial Stability,” European Systemic Risk Board, June.

European Commission (2019), “Guidelines on Reporting Climate-related Information,” Directorate-General for Financial Stability, Financial Ser-vices and Capital Markets Union.

Fay, M., S. Hallegatte, A. Vogt-Schilb, J. Rozenberg, U. Narloch, and T. Kerr (2015), “Decarbonizing Development: Three Steps to a Zero-Car-bon Future,” World Bank Group, Washington, D.C.

FSB (2015), “Proposal for a Disclosure Task Force on Climate-related Risks,” Financial Stability Board, https://www.fsb-tcfd.org/wp-content/uploads/2016/01/FSB_Disclosure-task-force-on-climaterelated-risks.pdf.

Fischer, Y. (2019), “Global Warming: Does the ECB Mandate Legally Au-thorise a ‘Green Monetary Policy’?” Business & Law Research Centre, 17.

Gabor, D. (2021), “Revolution Without Revolutionaries: Interrogating the Return of Monetary Financing,” Transformative Responses to the crisis, Finanzwende, Heinrich-Boll-Foundation.

Grippa, P., J. Schmittmann, and F. Suntheim (2019), “Climate Change and Financial Risk: Central Banks and Financial Regulators are Starting to Factor in Climate Change,” Finance & Development 56, 4.

Hepburn, C., N. Stern, and J. E. Stiglitz (2020), “Carbon Pricing,” Special issue in the European Economic Review 127.

High-Level Commission on Carbon Prices (2017), Report of the High­Level Commission on Carbon Prices, World Bank, Washington, D.C.

Honohan, P. (2019), “Should Monetary Policy Take Inequality and Cli-mate Change into Account?” Peterson Institute for International Eco-nomics, Working Paper 19–18.

IMF (2019), Fiscal Monitor, International Monetary Fund, Washington, D.C, October.

IMF (2020a), “Mitigating Climate Change,” World Economic Outlook Re­port, Chapter 3, October.

IMF (2020b), Global Financial Stability Report, Chapter 5, April.

IMF (2021), “The Economics of Biodiversity. Conversation with Professor Sir Partha Dasgupta and Kristalina Georgieva,” https://www.imf.org/en/News/Seminars/Conferences/2021/01/21/the-economics-of-biodiversity.

IPBES (2020), “Workshop Report on Biodiversity and Pandemics of the Intergovernmental Platform on Biodiversity and Ecosystem Services,” IPBES Secretariat, Bonn.

IPCC (2018), Global Warming of 1.5°C, Intergovernmental Panel on Cli-mate Change, Switzerland.

IPCC (2019), Special Report on Climate Change and Land, Intergovern-mental Panel on Climate Change, Switzerland.

Johnson, C. K., P. L. Hitchens, P. S. Pandit, J. Rushmore, T. S. Evans, C. C. W. Young, and M. M. Doyle (2020), “Global Shifts in Mammalian Population Trends Reveal Key Predictors of Virus Spillover Risk,” Pro­ceedings of the Royal Society B.

Kandlikar, M., J. Risbey, and S. Dessai. (2005), “Representing and Com-municating Deep Uncertainty in Climate Change Assessments,” Comptes Rendus Geoscience 337, 443–55.

Kapp, K. W. (1950), The Social Costs of Private Enterprise, Harvard Uni-versity Press, Cambridge.

Kedward, K., J. Ryan-Collins, and H. Chenet (2020), “Managing Nature- related Financial Risks: A Precautionary Policy Approach for Central Banks and Financial Supervisors,” UCL Institute for Innovation and Pub-lic Purpose Working Paper 2020-09.

Keynes, J. M. (1936), The General Theory of Employment, Interest and Money, 2013 Edition, Cambridge University Press.

Knight, F. H. (1921), Risk, Uncertainty and Profit, University of Illinois at Urbana-Champaign’s Academy for Entrepreneurial Leadership Historical Research Reference in Entrepreneurship.

Krogstrup, S. and W. Oman (2019), “Macroeconomic and Financial Pol-icies for Climate Change Mitigation: A Review of the Literature,” IMF Working Paper 19/185, International Monetary Fund, Washington, D.C.

Lagarde, C. (2020), Introductory remarks by Christine Lagarde, Pres-ident of the ECB, at the Franco-German Parliamentary Assembly, September 21.

Lenton, T. M., J. Rockstrom, O. Gaffney, S. Rahmstorf, K. Richardson, W. Steffen, and H. J. Schnellhuber (2019), “Climate Tipping Points — Too Risky to Bet Against,” Nature 575, 592–595.

Lipsey R. G., and K. Lancaster (1956), “The General Theory of Second Best,” Review of Economic Studies 24(1), 11–32.

McGlade, C., and P. Ekins (2015), “The Geographical Distribution of Fos-sil Fuels Unused When Limiting Global Warming to 2°C,” Nature 517, 187–190.

Mercure, J. F., H. Pollitt, J. E. Viñuales, N. R. Edwards, P. B. Holden, U. Chewpreecha, P. Salas, I. Sognnaes, A. Lam, and F. Knobloch (2018), “Macroeconomic Impact of Stranded Fossil Fuel Assets,” Nature Climate Change 8(7), 588–93.

Merton, R. (1995), “A Functional Perspective of Financial Intermedia-tion,” Financial Management 24(2), 23–41.

NGFS (2019), “NGFS First Comprehensive Report. A Call for Action – Cli-mate Change as a Source of Financial Risk,” Network of Central Banks and Supervisors for Greening the Financial System.

NGFS (2020a), “A Status Report on Financial Institutions’ Experiences from Working with Green, Non Green and Brown Financial Assets and a Potential Risk Differential,” Network of Central Banks and Supervisors for Greening the Financial System.

NGFS (2020b), “Survey on Monetary Policy Operations and Climate Change: Key Lessons for Further Analysis,” Network of Central Banks and Supervisors for Greening the Financial System.

NGFS (2020c), “Guide to climate scenario analysis for central banks and supervisors,” Network of Central Banks and Supervisors for Greening the Financial System.

OECD (2015), Aligning Policies for a Low­carbon Economy, OECD Publish-ing, Paris.

Oustry, A., B. Erkan, R. Svartzman, and P.-F. Weber (2020), “Climate-re-lated Risks and Central Banks’ Collateral Policy: A Methodological Exper-iment,” Banque de France Working Paper No. 790.

Parker, M. (2018), “The Impact of Disasters on Inflation,” Economics of Disasters and Climate Change 2, 21–48.

Pigato, M., ed. (2019), Fiscal Policies for Development and Climate Action, World Bank Group, Washington, D.C.

Pisani-Ferry, J. (2021), “Central Banking’s Brave New World,” Project Syndicate, February 23.

CONTENT

11CESifo Forum 3 / 2021 May Volume 22

FOCUS

Rafaty, R., G. Dolphin, and F. Pretis (2020), “Carbon Pricing and the Elasticity of CO2 Emissions,” Cambridge Working Papers in Economics, 20116.

Rockstrom, J., W. Steffen, K. Noone, Å. Persson, F. S. Chapin III, E. F. Lambin, T. M. Lenton, M. Scheffer, C. Folke, H . J. Schellnhuber, and B. Nykvist (2009), “A Safe Operating Space for Humanity,” Nature 461, 472–475.

Schellekens, G. and J. van Toor (2019), “Values at Risk? Sustainability Risks and Goals in the Dutch Financial Sector,” De Nederlandsche Bank.

Schnabel, I. (2020), “When Markets Fail – the Need for Collective Action in Tackling Climate Change,” Speech at the European Sustainable Fi-nance Summit, Frankfurt am Main, September 28.

Schnabel, I. (2021), “From Green Neglect to Green Dominance?” Inter-vention at the “Greening Monetary Policy – Central Banking and Climate Change,” online seminar, organized as part of the “Cleveland Fed Con-versations on Central Banking,” March 3.

Schoenmaker, D. (2021), “Greening Monetary Policy,” Climate Policy.

Slater, T., I. R. Lawrence, I. N. Otosaka, A. Shepherd, N. Gourmelen, L. Jakob, P. Tepes, L. Gilbert, and P. Nienow (2021), “Review Article: Earth’s Ice Imbalance,” The Cryosphere 15, 233–246.

Steffen, W., J. Rockstrom, K. Richardson, T. M. Lenton, et.al. (2018), “Tra-jectories of the Earth System in the Anthropocene,” Proceedings of the National Academy of Sciences of the United States of America 115(33), 8252–8259.

Stern, N. (2007), The Economics of Climate Change: The Stern Review, Cambridge University Press, Cambridge.

Stern, N. (2015), Why Are We Waiting? The Logic, Urgency, and Promise of Tackling Climate Change, The MIT Press, Cambridge.

Stern, N. and J. E. Stiglitz (2021), “The Social Cost of Carbon, Risk, Distribution, Market Failures: An Alternative Approach,” NBER Working Paper No. 28472.

Svartzman, R., P. Bolton, M. Després, L. A. Pereira da Silva, and F. Sa-mama (2020), “Central Banks, Financial Stability and Policy Coordina-tion in the Age of Climate Uncertainty: A Three-layered Analytical and Operational Framework,” Climate Policy.

Svartzman, R., and J. Althouse (2020), “Greening the International Mon-etary System? Not Without Addressing the Political Ecology of Global Imbalances,” Review of International Political Economy.

Tänzler, D., and N. Gordon (2020), “The New Geopolitics of a Decarbon-izing World,” Wilson Center.

UNFCCC (2016), Paris Agreement, United Nations Framework Convention on Climate Change, New York City, https://unfccc.int/sites/default/files/english_paris_agreement.pdf.

Van Tilburg, R. and A. Simic (2021), “Every Avenue Available: Lessons from Monetary History for Tackling Climate Change,” Sustainable Fi-nance Lab.

Van ‘t Klooster, J., and R. van Tilburg (2020), “Targeting a sustainable recovery with Green TLTROs,” Positive Money and Sustainable Finance Lab.

Vermeulen, R., E. Schets, M. Lohuis, B. Kolbl, D.-J. Jansen, and W. Hee-ringa (2018), “An Energy Transition Risk Stress Test for the Financial System of the Netherlands,” De Nederlandsche Bank Occasional Studies 16–7.

Villeroy de Galhau, F. (2015), “Climate Change—The Financial Sector and Pathways to 2°C,” Speech on November 30, http://www.bis.org/review/r151229f.pdf.

Villeroy de Galhau, F. (2021), “The Role of Central Banks in the Greening of the Economy,” Speech on February 11, https://www.bis.org/review/r210211g.htm.

Weidmann, J. (2020), “Bundesbank Chief: How Central Banks Should Address Climate Change,” November 18, https://www.ft.com/content/ed270eb2-e5f9-4a2a-8987-41df4eb67418.

Weidmann, J. (2021), “What Role Should Central Banks Play in Com-bating Climate Change?” Remarks at the ILF Online-Conference “Green Banking and Green Central Banking: What are the Right Concepts?” Goethe University Frankfurt, 25 January.

Weitzman, M. L. (2009), “On Modeling and Interpreting the Economics of Catastrophic Climate Change,” Review of Economics and Statistics 91(1), 1–19.

Weitzman, M. L. (2012), “GHG Targets as Insurance Against Catastrophic Climate Damages,” Journal of Public Economic Theory 14(2), 221–244.

Zenghelis, D. and N. Stern (2016), “The Importance of Looking Forward to Manage Risks: Submission to the Task Force on Climate-Related Fi-nancial Disclosures,” Center for Climate Change Economics and Policy and Grantham Research Institute on Climate Change and the Environ-ment Policy Paper.

CONTENT

12 CESifo Forum 3 / 2021 May Volume 22

FOCUS

Over the past decade, environmental, social and gov-ernance (ESG) considerations have been among the most important factors responsible for shifting the axes of the financial industry. What began as a passive form of ‘exclusion’ based investing (Heinkel, Kraus, and Zechner 2001; Hong and Kacperczyk 2009), has morphed into an array of investment strategies, in-cluding proactive shareholder ESG engagement (Dyck, Lins, Roth, and Wagner 2019; Hoepner, Oikonomou, Sautner, Starks, and Zhou 2020; Barko, Cremers, and Renneboog 2018). According to the US SIF Founda-tion, the aggregate amount of assets under manage-ment incorporating ESG in one form or another has grown over fivefold since 2010, standing at USD 15 trillion in the US alone as of 2020. Hartzmark and Sussman (2019) and Ceccarelli, Ramelli, and Wagner (2020) show that investors are indeed attracted to ESG investments.

A natural and important question to ask is: What has driven this growth? More importantly, what explains such investor behavior? Are investors attracted to ESG factors because they are superior in terms of their risk-return tradeoffs or their use-fulness as a hedging device, in other words for fi-nancial considerations? Or is there a distinct ‘taste’ for ESG that has developed recently, giving in- vestors non-pecuniary utility such as the moral sa-tisfaction of having made an environmental and social impact? Do these different types of demand for ESG investing have distinct implications for its future? This is one of the biggest financial ques-tions among practitioners, and academics have begun to provide answers. In this article we review the recent burgeoning academic literature in finan-cial eco nomics in an attempt to understand ESG investments and the investor preferences driving such investments.

ESG INVESTMENTS AND HEDGING DOWNSIDE RISK

What are the potential financial incentives of ESG investing? Undoubtedly, one of the greatest motiva-tions is to hedge climate change related long-term risk, which has indeed been widely documented as a significant risk factor, as perceived particularly by large and sophisticated institutional investors (Krue-ger, Sautner, and Starks 2020). As such, institutional investors have pushed for increased disclosure of such risk exposures by the companies they invest in (Ilhan, Krueger, Sautner, and Starks 2020).

An important strand of academic research has focused on how to theoretically conceptualize cli-mate risk and the implications of climate risk on as-set prices (Barnett 2020; Barnett, Brock, and Hansen 2020). A large amount of work has explored whether this risk is priced in asset markets, such as equities, options, or real estate, on the basis of actual time-se-ries and cross-sectional data, across multiple asset classes, and under various climate risk related circum-stances, such as extreme weather events or sea level rise (Bolton and Kacperczyk 2020; Baldauf, Garlappi, and Yannelis 2020; Bernstein, Gustafson, and Lewis 2019; Kruttli, Roth Tran, and Watugala 2021). The gen-eral consensus is that climate is indeed a significant source of risk, reflected in the return premia on assets with high climate risk exposure. It is for this reason that investors are interested in strategies to hedge against this risk (Engle, Giglio, Kelly, Lee, and Stroebel 2020; Giglio, Kelly, and Stroebel 2020).

The concern of ESG investors regarding down-side risk goes beyond climate risk. In fact, the litera-ture on corporate social responsibility (CSR) argues and demonstrates that corporate investment in CSR is a useful hedge against downside risk in general (Lins, Servaes, and Tamayo 2017; Albuquerque, Koski-

nen, and Zhang 2019). For example, when a company suffers a reputational or economic

shock, prior investments in CSR may ensure customer and employee loyalty or signal differentiation against competitors, protect-ing the firm against such shocks.

The Covid-19 economic crisis that has impacted businesses worldwide is a perfect example of such a shock. Several recent stud-ies show that the stock prices of firms that had high ESG performance suffered much less during the market crash that followed the onset of Covid-19 (Albuquerque, Koskinen, Yang, and Zhang 2020; Ding, Levine, Lin, and Xie 2020; Pastor and Vorsatz 2020).

Robin Döttling and Sehoon Kim

ESG Investments and Investors’ Preferences

is Assistant Professor in the De-partment of Finance at Erasmus University Rotterdam. His re-search focuses on financial inter-mediation, corporate finance and macroeconomics.

is Assistant Professor in the De-partment of Finance, Insurance and Real Estate at University of Florida. His research interests include corporate finance, sus-tainability, and financial markets among others.

Robin Döttling Sehoon Kim

CONTENT

13CESifo Forum 3 / 2021 May Volume 22

FOCUS

Consistent with such hedging benefits of ESG, Hoepner et al. (2020) show that shareholder engage-ment by institutional investors on ESG-related issues indeed substantially reduces the target firm’s down-side risk exposure as measured by value-at-risk.

NON-PECUNIARY MOTIVATIONS FOR ESG INVESTING

Another driver behind the exponential growth in ESG investments may be that investors obtain non-pe-cuniary utility from aligning their investments with their social preferences (‘warm glow’ in the words of Andreoni 1989). While it is costly for companies to internalize environmental and social externalities in their investments, Hart and Zingales (2017) argue that if shareholders have pro-social preferences, firms should pursue policies that maximize a shareholder ‘welfare’ that incorporates these pro-social prefer-ences, rather than purely focusing on market value as propagated by Friedman (1970).

An example of such an approach to explicitly incorporating a non-pecuniary utility for pro-social investments in the investor’s utility function is Fama and French (2007), in which the authors model as-sets as consumption goods for which investors have tastes. They show that investors with tastes for as-sets that resemble tastes for socially responsible in-vesting earn negative alphas in equilibrium. Baker, Bergstresser, Serafeim, and Wurgler (2018) present a similar CAPM framework, in which investors obtain non-pecuniary utility from holding green bonds and use this framework to rationalize empirical evidence that green bonds are priced at a premium relative to regular bonds.

A large body of research has emerged, providing various types of empirical evidence that is consistent with such non-pecuniary motives for ESG investing. For example, ESG fund flows are less volatile and also less sensitive to negative returns, in marked contrast to the conventional wisdom that fund flows tend to be highly sensitive to performance (Bollen 2007; Renneboog, Ter Horst, and Zhang 2011). Riedl and Smeets (2017) combine mutual fund holdings data with experimental and survey evidence, showing that investors who behave more socially in a trust game experiment and donate more to charity also hold more socially responsible equity funds, which is consistent with pro-social preferences driving ESG investments. Moreover, their survey evidence in- dicates that these investors invest in ESG funds even though they expect those funds to underperform, indicating that investors are willing to give up return to align their investments with social preferences. Experimental evidence in Humphrey, Kogan, Sagi, and Starks (2020) shows that individuals adjust their investment strategy if investments have negative externalities on charities they care about. Interest-ingly, their experimental design allows the effect

of social preferences to be identified, because pay-offs are set to be identical regardless of the social setting, effectively controlling for risk and return consid erations. Further evidence of social preferences driving investment behavior can be found in Bauer, Ruof and Smeets (2019), who find that Dutch pen-sion plan participants prefer their pension fund to focus on sustainable development goals, as well as in survey evidence by Brodback, Guenster and Mezger (2019).

COMBINATION OF BOTH ESG DEMAND FACTORS

The existing evidence indicates that the hedging-de-mand explanation and pro-social motives both play a role in the recent popularity of ESG investing. Inves-tors who care about climate risk and other downside risk and investors who have non-pecuniary motives can coexist in the market. Indeed, such investor het-erogeneities, or shifts in investor ‘taste’ for ESG, are at the center of recent theoretical frameworks of ESG investing.

For example, Pastor, Stambaugh and Taylor (2020) incorporate tastes for ‘green’ investments in investor’s preferences and deliver additional in-sights relative to the previous literature. First, their model shows that even absent non-pecuniary ben-efits ESG investments may be priced at a premium because they provide a hedge against climate risks. Intuitively, ‘brown’ assets perform poorly when cli-mate risks materialize and environmental regulation tightens. Green assets hedge against such risks, war-ranting a lower expected return. Second, the model shows that while green assets should have a lower expected return in the long-run, they may outper-form brown assets in the short run if there are pos-itive shocks to investor sustainability preferences. Alternatively, Pedersen, Fitzgibbons, and Pomorski (2020) show that assets with higher ESG scores may have higher expected returns if many investors are unaware of ESG scores and fail to bid up the price of high ESG assets but may have lower expected re-turns if many investors have non-pecuniary utility for ESG. Oehmke and Opp (2020) show how socially responsible investors can increase corporate green investment by alleviating financing frictions. Socially responsible investors are intuitively willing to accept a lower return on their financing terms if it induces firms to shift from polluting to non-polluting invest-ments, lowering financing costs and alleviating fi-nancial constraints.

COVID-19 AND RETAIL MUTUAL FUND INVESTOR ESG PREFERENCES

An open question in the literature is how ESG in-vestment flows of different investor classes respond to an adverse economic shock. In our recent study (Dottling and Kim 2021), we answer this question by

CONTENT

14 CESifo Forum 3 / 2021 May Volume 22

FOCUS

investigating the fragility (or ‘sustainability’) of ESG demand by retail mutual fund investors in the face of the great economic shock brought about by Covid-19. Our evidence shows that retail demand for ESG in-vestment dropped significantly in response to the Covid-19 shock, consistent with the demand driven by pro-social preferences that become less affordable to pursue under economic distress. In stark contrast, we document resilient ESG demand by institutional investors, who often invest in ESG as a hedge against downside risk.

The economic crisis that has followed the out-break of the novel coronavirus provides an ideal setting for studying the impact of a significant and unexpected economic shock on ESG investment pref-erences. First, it has triggered the first major crisis originating in the real economy seen during this pe-riod of growth in sustainable investing. Second, the root cause of the economic shock was unrelated to economic preconditions and heterogeneous across regions, helping to establish a causal link between economic distress and ESG demand. Third, the con-tinuous deterioration in economic conditions con-trasts starkly with the initial stock market crash and subsequent post-stimulus rebound, helping distin-guish shifts in ESG demand from simple valuation effects.

Using this shock, we can study the response of investments by retail investors in mutual funds with high Morningstar sustainability ratings as a measure of revealed preference for ESG. Retail sustainable fund flows are a suitable measure, given that retail inves-tors are known to actively reallocate capital across funds in response to sentiment and preference shifts. Retail investors are also economically important, rep-resenting over 60 percent of total net mutual fund assets.

Based on Morningstar data covering US domi-ciled open-end equity mutual funds and their weekly retail fund flows as well as sustainability ratings, we find that retail investor demand for ESG invest-ments weakens substantially under the economic

stress imposed by Covid-19. Funds with the highest (five-globe) Morningstar sustainability ratings that received higher than average flows prior to the cri-sis experience a sharper decline in flows after the onset of the pandemic-induced economic downturn, dropping to the level of funds with low sustainabil-ity ratings.

This key result is illustrated in Figure 1, where we plot weekly average retail fund flows over the sample period from January 4 to April 25, with different sus-tainability ratings. Our estimates imply that weekly net flows into highly rated funds as a fraction of be-ginning-of-week net assets dropped 0.2 percentage points further from pre-Covid levels during the crisis, compared to average funds.

Moreover, this shift not only manifests early dur-ing the market crash weeks between February 22 and March 21, but also persists between March 28 and April 25, when the stock market rebounded dramati-cally after the US stimulus package was announced, while the economy continued to deteriorate.

INSTITUTIONAL FLOWS CONTINUE INTO SUSTAINABLE FUNDS

Our interpretation is that investor demand for sus-tainable investments is sensitive to economic con-ditions, consistent with retail investor ESG demand driven by pro-social motives, the pursuit of which becomes less affordable during economic distress (akin to a ‘luxury good’). To test this, we conduct a comparison of retail and institutional fund flows. Distinguishing between mutual fund investments by retail and institutional investors provides important clues to help explain the retail flow responses. For example, institutional investors are much less finan-cially constrained and face higher minimum invest-ment requirements (e.g., USD 200,000 or more). More-over, institutional investors are likely to have strong explicit ESG mandates and perform ESG shareholder engagements as part of their core strategies to hedge against risks.

Given these differences, one would expect eco-nomically distressed retail investors to cut spending on ESG more sharply in bad times than deep-pocketed and committed institutions do, if indeed ESG invest-ment is driven by pro-social preferences and viewed as a ‘non-essential good’. Consistent with this idea, we find that institutional flows do not disproportion-ately decline for high sustainability funds in response to the Covid-19 crisis.

This is illustrated in Figure 2, where we plot weekly institutional fund flows. In fact, we find that institutional flows drop sharply – mainly for low sus-tainability funds – but only temporarily during the early market crash period. This further indicates that our main result on retail flows is consistent with non-pecuniary preferences for ESG, a perceived ‘lux-

0.2

0.0

–0.8

–0.6

–0.4

–0.2

Net flow (%)

Source: Authors‘ calculations.

Jan 4 Jan 18 Feb 1 Feb 15 Feb 29 Mar 14 Mar 28 Apr 11 Apr 25

High Sustainability Average Sustainability Low Sustainability

Weekly Retail Fund Flows by Sustainability Rating

© ifo Institute

Figure 1

CONTENT

15CESifo Forum 3 / 2021 May Volume 22

FOCUS

ury good’ that retail investors can no longer afford under economic distress.

INTERNATIONAL EVIDENCE

To corroborate our interpretation that the drop in retail ESG demand is driven by economic distress, we extend our U.S.-based evidence to an international sample to allow us to exploit heterogeneity in the severity of the Covid-19 economic shock across regions. We find that the drop in retail investor ESG demand also occurred in Europe as well as in a broad sample that includes all funds in the Morningstar database globally. Interestingly, the drop in ESG demand is significantly larger in regions that ex-perienced stricter lockdowns or less economic support as measured by the Oxford Covid-19 Gov-ernment Response Tracker and regions that had lower GDP growth in the first two quarters of 2020. This heterogeneity suggests that the drop in ESG in-vestment demand is driven by the economic distress brought about by the Covid-19 crisis and further cor-roborates our interpretation of such demand driven by non-pecuniary utility from ESG investments.

SHIFT IN INTERESTS FROM SUSTAINABILITY TO FINANCIAL AND ECONOMIC ISSUES

Our interpretation is also supported by evidence of Google search traffic shifting away from topics related to sustainability or ESG toward issues con-cerning the economy (see Figure 3). We further ex-clude potential alternative explanations based on conventional factors known to affect fund flows, such as fund style, age, size, expenses, past returns and flows, star ratings, time trends, buying-losers or selling-winners behavior, or changes in risk toler-ance. Overall, we find our evidence consistent with non-financial incentives for sustainable investments by retail investors that are adversely impacted by a large-scale economic crisis.

IMPLICATIONS FOR ESG INVESTING

A long-term implication of our findings is a fragility and cyclicality in ESG demand stemming from retail investors. Under prolonged economic distress, this means that there could potentially be a broader shift in investor preferences – retail demand may also af-fect the institutional push for ESG and weaken the intensity and influence of institutional ESG engage-ment. Thus, to make sustainable investing ‘sustain-able’, individual investors must clearly understand any tangible financial and economic value that may arise from ESG investing. Leaning on social signals and preferences to attract ESG demand will only sustain such investments for so long, mostly when times are good and the societal need for ESG investments is at its lowest.

REFERENCES Albuquerque, R., Y. Koskinen, S. Yang, C. Zhang (2020), “Resiliency of En-vironmental and Social Stocks: An Analysis of the Exogenous COVID-19 Market Crash,” Review of Corporate Finance Studies 9(3), 593–621.

Albuquerque, R., Y. Koskinen, and C. Zhang (2019), “Corporate Social Responsibility and Firm Risk: Theory and Empirical Evidence,” Manage­ment Science 65(10), 4451–4469.

Andreoni, J. (1989), “Giving With Impure Altruism: Applications to Charity and Ricardian Equivalence,” Journal of Political Economy 97(6), 1447–1458.

Baker, M., D. Bergstresser, G. Serafeim, and J. Wurgler (2018), “Financ-ing the Response to Climate Change: The Pricing and Ownership of US Green Bonds,” National Bureau of Economic Research, Working Paper 25194.

−1.0

–0.5

0.0

0.5

1.0Net flow (%)

Jan 4 Jan 18 Feb 1 Feb 15 Feb 29 Mar 14 Mar 28 Apr 11 Apr 25Source: Authors‘ calculations.

High Sustainability Average Sustainability Low Sustainability

Weekly Institutional Fund Flows by Sustainability Rating

© ifo Institute

Figure 2

0

20

40

60

80

100

January February March April May

Coronavirus Global warming ESG SustainabilitySearch traffic

Internet Search Traffic on Sustainability-Related and Other Issues 2020

0

20

40

60

80

100

January February March April May

Coronavirus Stock market Furlough Financial crisisSearch traffic

© ifo InstituteSource: Authors' illustration based on Morningstar data.

Figure 3

CONTENT

16 CESifo Forum 3 / 2021 May Volume 22

FOCUS

Baldauf, M., G. Lorenzo, and Y. Constantine (2020), “Does Climate Change Affect Real Estate Prices? Only If You Believe In It,” The Review of Financial Studies 33(3), 1256–1295.

Barnett, M., W. Brock, and L. P. Hansen (2020), “Pricing Uncertainty In-duced by Climate Change,” Review of Financial Studies 33(3), 1024–1066.

Barko, T., M. Cremers, and L. Renneboog (2018), “Shareholder Engage-ment on Environmental, Social, and Governance Performance,” Working Paper.

Bauer, R., T. Ruof, and P. Smeets (2020), “Get Real! Individuals Prefer More Sustainable Investments,” Working paper.

Bernstein, A., M. T. Gustafson, and R. Lewis (2019), “Disaster on the Ho-rizon: The Price Effect of Sea Level Rise,” Journal of Financial Econom­ics 134(2), 253–272.

Bollen, N. PB (2007), “Mutual Fund Attributes and Investor Behavior,” Journal of Financial and Quantitative Analysis 42(03), 683–708.

Bolton, P. and M. Kacperczyk (2020), “Do Investors Care About Carbon Risk?,” National Bureau of Economic Research, Working Paper 26968.

Brodback, D., N. Guenster, and D. Mezger (2019), “Altruism and Egoism in Investment Decisions,” Review of Financial Economics, 37(1), 118–148.

Ceccarelli, M., S. Ramelli, and A. F. Wagner (2021), “Low-carbon Mutual Funds,” Swiss Finance Institute Research Paper 19–13.

Ding, W., R. Levine, C. Lin, and W. Xie (2021), “Corporate Immunity to the COVID-19 Pandemic,” Journal of Financial Economics.

Dottling, R. and S. Kim (2020), “Sustainability Preferences Under Stress: Evidence from Mutual Fund Flows During COVID-19,” Working Paper.

Dyck, A., K. V. Lins, L. Roth, and H. F. Wagner (2019), “Do Institutional Investors Drive Corporate Social Responsibility? International Evidence,” Journal of Financial Economics 131(3), 693–714.

Engle, R. F., S. Giglio, B. Kelly, H. Lee, and J. Stroebel (2020), “Hedging Climate Change News,” Review of Financial Studies 33(3), 1184–1216.

Fama, E. F. and K. R. French (2007), “Disagreement, Tastes, and Asset Prices,” Journal of Financial Economics 83(3), 667–689.

Friedman, M (1970), A Friedman Doctrine: The Social Responsibility of Business is to Increase Its Profits, The New York Times Magazine, Sep-tember 13.

Giglio, S., B. T. Kelly, and J. Stroebel (2020), “Climate Finance,” National Bureau of Economic Research, Working Paper 28226.

Gregory, R. P (2021), “Climate Disasters, Carbon Dioxide, and Financial Fundamentals,” Quarterly Review of Economics and Finance 79, 45–58.

Hart, O. and L. Zingales (2017), “Serving Shareholders Doesn’t Mean Putting Profit Above All Else,” Harvard Business Review 12, 2–6.

Hartzmark, S. M. and A. B. Sussman (2019), “Do Investors Value Sustain-ability? A Natural Experiment Examining Ranking and Fund Flows,” Jour­nal of Finance 74(6), 2789–2837.

Heinkel, R., A. Kraus, and J. Zechner (2001), “The Effect of Green In-vestment on Corporate Behavior,” Journal of Financial and Quantitative Analysis 36(4), 431–449.

Hoepner, A. G., I. Oikonomou, Z. Sautner, L. T. Starks, and X. Zhou (2020), “ESG Shareholder Engagement and Downside Risk,” Working Paper.

Hong, H. and M. Kacperczyk (2009), “The Price of Sin: The Effects of So-cial Norms on Markets,” Journal of Financial Economics 93(1), 15–36.

Humphrey, J., S. Kogan, J. S. Sagi, and L. T. Starks (2020), “The Asym-metry in Responsible Investing Preferences,” Working Paper.

Krueger, P., Z. Sautner, and L. T. Starks (2020), “The Importance of Climate Risks for Institutional Investors,” The Review of Financial Stud­ies 33(3), 1067–1111.

Ilhan, E., P. Krueger, Z. Sautner, L. T. Starks (2020), “Climate Risk Disclo-sure and Institutional Investors”, Swiss Finance Institute Research Paper, 19–66.

Lins, K. V., H. Servaes, and A. Tamayo (2017), “Social Capital, Trust, and Firm Performance: The Value of Corporate Social Responsibility During the Financial Crisis,” Journal of Finance 72(4), 1785–1824.

Oehmke, M. and M. M. Opp (2020), “A Theory of Socially Responsible Investment,” Working Paper.

Pástor, Ľ., R. F. Stambaugh, and L. A. Taylor (2020), “Sustainable Invest-ing in Equilibrium,” Journal of Financial Economics.

Pástor, Ľ. and M. B. Vorsatz (2020), “Mutual Fund Performance and Flows During the COVID-19 Crisis,” Review of Asset Pricing Studies 10(4), 791–833.

Pedersen, L. H., S. Fitzgibbons, and L. Pomorski (2020), “Responsible Investing: The ESG-efficient Frontier,” Journal of Financial Economics.

Renneboog, L., J. Ter Horst, and C. Zhang (2011), “Is Ethical Money Financially Smart? Nonfinancial Attributes and Money Flows of So-cially Responsible Investment Funds,” Journal of Financial Intermedia­tion 20(4), 562–588.

Riedl, A. and P. Smeets (2017), “Why Do Investors Hold Socially Respon-sible Mutual Funds?,” Journal of Finance 72(6), 2505–2550.

CONTENT

17CESifo Forum 3 / 2021 May Volume 22

FOCUS

Our society faces several grand challenges, such as climate change, poverty, inequality, and global health. As public governance is unlikely to be sufficient to address these challenges, a large responsibility rests on the shoulders of the private sector. Accordingly, it is important to understand how and to what extent companies (and investors) can grow and sustain their organizations over time while strengthening—rather than undermining—the very system in which they op-erate, as they can play a critical role in addressing these societal grand challenges.

THE GOVERNANCE CHALLENGE: A LACK OF LONG-TERM ORIENTATION AND PRIVATE INCENTIVES

In a series of studies, I explore the question of whether and how companies’ social and environmental en-gagement—in short, their corporate social responsi-bility (CSR)―contributes to their competitiveness. As a whole, the evidence indicates that CSR can benefit companies along several dimensions that are core to their competitiveness and enhance their financial per-formance. These dimensions include i) the pursuit of innovation (Flammer and Kacperczyk 2016), ii) the re-tention of knowledge (Flammer and Kacperczyk 2019), iii) employee engagement (Flammer and Luo 2016), iv) differentiation from competitors on the product market (Flammer 2015a), v) differentiation from com-petitors on the market for government procurement contracts (Flammer 2018), and vi) resilience in times of economic crisis (Flammer and Ioannou 2020). As a result, it is perhaps not surprising that CSR is found to positively affect shareholders’ perceptions and share-holder returns (Flammer 2013; 2015b).

Taken together, these studies indicate that com-panies’ social and environmental practices can en-hance a firm’s competitiveness and financial per-formance. As such, this suggests that a firm’s CSR should be core to its corporate governance and an integral part of its strategy. In other words―and re-ferring to the ‘ESG’ (environmental, social, and cor-porate governance) acronym often used among prac-titioners―this suggests that the ‘E’ and ‘S’ are not separate but rather an integral part of ‘G’ and hence should be more central to firms’ strategy than it is often considered to be. By integrating ‘E’ and ‘S’ into their decision making, companies can enhance not only their profitability and ability to create long-term value, but also exert a positive impact on the natural environment and society at large.

Given the potential benefits of ‘E’ and ‘S’, it may seem puzzling that companies do not engage more extensively in CSR. My research identifies two po-

tential reasons: i) short-termism, and ii) the lack of private incentives. In two related studies, I exam-ine how two governance mecha-nisms—namely the provision of long-term executive compensa-tion and CSR-based executive compensation—can alleviate these challenges and induce a longer-term orientation that is conducive to more sustainable business practices. These studies are summarized in the following.

EXECUTIVE COMPENSATION TIED TO LONG-TERM FINANCIAL PERFORMANCE

In the article “Does a Long-term Orientation Create Value? Evidence from a Regression Discontinuity” (Flammer and Bansal 2017), Pratima Bansal and I investigate how the pro-vision of long-term executive compensation—which aims to increase managers’ time horizons—impacts firm value and corporate strategy.

Specifically, we identify a new form of agency conflict, which we refer to as a time-based agency conflict—that is, managers’ and shareholders’ time preferences are potentially misaligned and, as a re-sult, managers may adopt short-sighted strategies that need not be in the shareholders’ best interests. This agency conflict arises if managers are overly concerned about ‘looking good in the short run.’ In particular, career concerns, executive compensation based on short-term performance, and the pressure to meet or exceed analysts’ earnings forecasts may lead managers to favor (inferior) short-term projects at the expense of (superior) long-term projects, thereby hurting the value of the firm. In an effort to mitigate managerial myopia and align managers’ interests with long-term value creation, boards of directors may pro-vide long-term incentives. To examine whether pro-viding such incentives is effective in fostering more sustainable business practices and improving firm value, we study the provision of long-term executive compensation (that is, compensation in the form of restricted stocks, restricted stock options, and long-term incentive plans).

In conducting the analysis, an important em-pirical challenge is that long-term compensation is not randomly assigned to companies, which makes it difficult to assess the causal impact of long-term compensation on performance (for example, it could be that companies that expect to be more profita-

Caroline Flammer

Short-Termism and Executive Compensation

is Associate Professor of Strategy & Innovation at Boston Univer-sity. She is also Dean’s Research Scholar and Fellow of the Susilo Institute. Her research interests include corporate governance, climate change, and corporate social responsibility among others.

Caroline Flammer

CONTENT

18 CESifo Forum 3 / 2021 May Volume 22

FOCUS

ble in the future are more likely to provide long-term compensation to their executives; this would induce a positive correlation between long-term compensa-tion and profitability, yet this correlation would not imply that long-term compensation causes higher profitability). To overcome this challenge, we ex-amine shareholder proposals advocating the use of long-term executive compensation that pass or fail by a small margin of votes at shareholder meetings. Intuitively, there should be no systematic difference between companies that marginally pass long-term compensation proposals with, say, 50.1 percent of the votes and companies that reject comparable propos-als with 49.9 percent of the votes. The passage of such ‘close call’ proposals is akin to a random assignment of long-term incentives to companies and therefore provides a quasi-experimental setting that lends itself to assess the causal effect of long-term incentives on firm performance. We then use a regression disconti-nuity design (RDD) to compare outcomes just above and below the majority threshold.

Using this RDD specification, we find that the passage of long-term compensation proposals leads to a positive stock market reaction. More precisely, on the day of the vote, a proposal that is marginally passed yields an abnormal return of 1.14 percent compared to a proposal that is marginally rejected. This evidence indicates that adopting a longer-term orientation is value-enhancing. We further examine the effect of passing long-term compensation propos-als on operating performance (return on assets, net profit margin, and sales growth). Regardless of the measure, we consistently find that operating perfor-mance increases in the long run. Interestingly, oper-ating performance decreases slightly in the short run (i.e., in the year following the vote), indicating that an increased long-term orientation may take some time to materialize into higher profits. Finally, we find that companies adopting long-term compensation pro-posals are more likely to increase their investments in long-term strategies such as CSR and innovation. This suggests that an increase in long-term orienta-tion benefits companies by fostering innovation and stakeholder relationships, allowing them to acquire intangible assets such as legitimacy, reputation, and trust.

EXECUTIVE COMPENSATION TIED TO SOCIAL AND ENVIRONMENTAL (CSR) PERFORMANCE

In an article entitled “Corporate Governance and the Rise of Integrating Corporate Social Responsibility Criteria in Executive Compensation: Effectiveness and Implications for Firm Outcomes” (Flammer, Hong, and Minor 2019), I further examine how executive com-pensation can help redirect managerial attention to long-term value creation. In this article, my co-au-thors, Bryan Hong and Dylan Minor, and I study the integration of CSR criteria in executive compensa-

tion—that is, the linking of executive compensation to social and environmental performance (e.g., CO2

emission targets, employee satisfaction targets, and compliance with ethical standards in developing coun-tries). Practitioners commonly refer to this incentive provision as ‘CSR contracting’ or ‘pay for social and environmental performance’ (as opposed to the tra-ditional ‘pay for (financial) performance’). The use of CSR contracting is a relatively recent practice in cor-porate governance and has received little attention in the academic literature.

To examine this new phenomenon, we construct a novel database that compiles information on CSR contracting from the compensation information that companies report in their proxy statements filed with the Securities and Exchange Commission (SEC). We start by documenting a series of stylized facts pertaining to CSR contracting. In particular, we ob-serve that the integration of CSR criteria in execu-tive compensation has become more prevalent over time (while 12 percent of the S&P 500 companies had adopted CSR contracting by 2004, this ratio increased to 37 percent by 2013), and is more common in emis-sion-intensive industries.

We then explore how CSR contracting affects firms’ outcomes. As in the case of executive com-pensation tied to long-term financial performance, we might expect that executive compensation tied to CSR criteria helps mitigate managerial myopia and improve corporate governance. This is indeed what we find. Specifically, we find that the adoption of CSR contracting leads to i) an increase in long-term ori-entation; ii) an increase in firm value; iii) an increase in social and environmental performance, especially with respect to less salient stakeholders such as the natural environment and communities; iv) a reduction in emissions; and v) an increase in green innovation. Moreover, we find that the results are stronger i) when companies specify the amount of CSR-based com-pensation involved (that is, when they are specific instead of vague), and ii) when the share of CSR-based compensation is larger. This suggests that CSR con-tracting is a more effective governance tool when it is substantive.

Overall, these findings indicate that CSR contract-ing enhances the governance of a company by incen-tivizing managers to adopt a longer time horizon and shift their attention towards stakeholders that are less salient in the short run, but financially material to the firm in the long run.

‘E’ AND ‘S’ AS INTEGRAL PARTS OF ‘G’

To conclude, the insights of these studies suggest that the ‘E’ and ‘S’ in ‘ESG’ are not separate elements but an integral part of ‘G’. Moreover, they suggest that corporate short-termism and the lack of private incen-tives hamper both engagement in more sustainable business practices and business success.

CONTENT

19CESifo Forum 3 / 2021 May Volume 22

FOCUS

Two potential remedies are the provision of long-term executive compensation and CSR-based executive compensation—both of these governance mechanisms induce managers to adopt a longer-term orientation that is conducive to long-term value cre-ation for the company and also benefits the natural environment and society at large. As such, these tools represent useful levers that boards of directors can pull to help address societal grand challenges in a way that also contributes to long-term value creation.

REFERENCES Flammer, C. (2013), “Corporate Social Responsibility and Shareholder Reaction: The Environmental Awareness of Investors,” Academy of Management Journal 56, 758–81.

Flammer, C. (2015a), “Does Product Market Competition Foster Corpo-rate Social Responsibility? Evidence from Trade Liberalization,” Strategic Management Journal 36, 1469–85.

Flammer, C. (2015b), “Does Corporate Social Responsibility Lead to Superior Financial Performance? A Regression Discontinuity Approach,” Management Science 61, 2549–68.

Flammer, C. (2018), “Competing for Government Procurement Contracts: The Role of Corporate Social Responsibility,” Strategic Management Journal 39, 1299–1324.

Flammer, C. and P. Bansal (2017), “Does a Long-Term Orientation Create Value? Evidence from a Regression Discontinuity,” Strategic Management Journal 38, 1827–47.

Flammer, C., B. Hong, D.B. Minor (2019), “Corporate Governance and the Rise of Integrating Corporate Social Responsibility Criteria in Executive Compensation: Effectiveness and Implications for Firm Outcomes,” Strategic Management Journal 40, 1097–1122.

Flammer, C. and I. Ioannou (2020), “Strategic Management during the Financial Crisis: How Firms Adjust their Strategic Investments in Re-sponse to Credit Market Disruptions,” Strategic Management Journal, forthcoming.

Flammer, C. and A. J. Kacperczyk (2016), “The Impact of Stakeholder Orientation on Innovation: Evidence from a Natural Experiment,” Man­agement Science 62, 1982–2001.

Flammer, C. and A. J. Kacperczyk (2019), “Corporate Social Responsi-bility as a Defense against Knowledge Spillovers: Evidence from the In evitable Disclosure Doctrine,” Strategic Management Journal 40, 1243–67.

Flammer, C. and J. Luo (2017), “Corporate Social Responsibility as an Employee Governance Tool: Evidence from a Quasi-Experiment,” Strate­gic Management Journal 38, 163–183.

CONTENT

20 CESifo Forum 3 / 2021 May Volume 22

FOCUS

The last decade has seen an expo-nential increase in corporate sus-

tainability activities and efforts by investors to use these activi-ties in their portfolio formation,

valuation, and stewardship ac-tivities. According to the UN Sus-

tainable Stock Exchanges, 44 ex-changes around the world have released ESG disclosure guidance for their listed companies. Nu-merous nonfinancial regulations, including one at the European Un-ion level, have promoted the dis-

closure of environmental, social, and governance (ESG) data (Gre-wal, Rield, and Serafeim 2019), and several not-for-profit or-

ganizations, such as the Global Reporting Initiative (GRI) and the Sustainability Accounting Stand-ards Board (SASB), have created disclosure standards that identify the ESG metrics that organizations should disclose. Figure 1 examines the assets under management of signatories of the Principles for Responsible Investing from 2013 to 2020, the value of total sustain-able debt issuances by companies,

and the volume of corporate ESG disclosures, as cal-culated by Bloomberg.

Figure 1 reports the changes in three ESG-related activities. The black line reports total assets under management (hundreds of billions USD) of signatories to the Principles for Responsible Investing. The grey

line reports total sustainable debt issuance (in bil-lions USD) (Bullard 2021). The blue line is the average ESG disclosure score for all companies, as calculated by Bloomberg.

One consequence of these activities has been the fundamentally different landscape surrounding what is measured and disclosed by companies, and what information is being used by investors to al-locate capital. On that latter point, the exponential increase in sustainability-linked debt represents an example of how traditional contractual terms — the payment of loan interest rate or bond coupons — is now becoming a function of ESG metrics. It might well be the case that in future, all financial instru-ments will be outcome-based, where their covenants or payment terms will be a function of ESG metrics. Contracting on ESG is also happening in executive compensation, where an increasing number of com-panies are tying incentives to ESG targets (Flammer, Hong, and Minor 2019).

Given that achieving environmental and social outcomes can be positive, neutral, or negative to fi-nancial performance, depending on the types of ac-tivities pursued, the quality of management, and the change in competitive dynamics, this shift in meas-urement and disclosure has also given rise to an ac-tive search for opportunities to better manage and deploy resources. More broadly, it has opened up the discussion about the role of the corporation in society and whether this extends beyond profit maximization (Mayer 2018). The discussion on the purpose of the corporation in society goes hand in hand with that on what evolution of the accounting system can provide the necessary accountability for the effective deploy-ment of firm resources and the impact of corporations on the environment and society.

A PARADIGM SHIFT

In his book “The Structure of Scientific Revolutions” (1962), Thomas Kuhn outlines how science progresses through periods of ‘normal science,’ whose conceptual continuity and cumulative progress are disrupted by revolutionary science and a paradigm shift. It is this paradigm shift that is currently taking place in the field of accounting.

Financial accounting, traditionally thought of as a system of measuring economic resources, liabili-ties and periodic performance based on changes over time, has emerged as one of the most important man-agement and accountability tools of modern history (Soll 2014). While a financial accounting system has

Ethan Rouen and George Serafeim

Impact-Weighted Financial Accounts: A Paradigm Shift

0

5

10

15

20

25

0

300

600

900

1200

1500

2013 2014 2015 2016 2017 2018 2019 2020

AUM of PRI signatories Total sustainable debt issuances Average ESG disclosure score

Note: Figure 1 reports the changes in three ESG-related activities. The black line reports total assets under management (AUM in hundreds of billions USD) of signatories to the Principles for Responsible Investing. The grey line reports total sustainable debt issuance (in billions USD). The blue line is the average ESG disclosure score for all companies.Source: Bullard (2021) and Bloomberg.

Changes in ESG Activites and Disclosures

Scaled dollars

© ifo Institute

ESG disclosure score

Figure 1

is Assistant Professor of Busi-ness Administration at Harvard Business School and the Faculty Co-Chair of the Impact-Weighted Accounts Project. His research revolves around how to measure the impact and value of human capital.

Ethan Rouen

is Charles M. Williams Professor of Business Administration at Harvard Business School and the Faculty Co-Chair of the Im-pact-Weighted Accounts Project. His research focuses on measur-ing, driving and communicating corporate performance and so-cial impact.

George Serafeim

CONTENT

21CESifo Forum 3 / 2021 May Volume 22

FOCUS

existed for thousands of years (Macve 2015), it is only in the last one hundred years that it has been widely practiced and adopted, since financial reporting has been mandated for firms with access to public capital markets and even firms that have remained private. Within this timeframe, accounting standards experi-enced significant changes with regard to leases, pen-sions, liabilities, and revenue recognition practices. However big these changes might seem, they hap-pened within the same intellectual paradigm. This paradigm measures assets, liabilities, revenues, and expenses, resulting in financial profit but ignoring any impacts that the organization has on employees, cus-tomers, or the environment. Within this paradigm, the addressees of accounting statements are the capital providers, and their information needs are satisfied by the current accounting framework.

Recently though, societal developments such as environmental degradation, increasing inequality, and the degradation of trust in institutions, have called into question whether financial accounting statements are fit for purpose in the 21st century (Eccles and Krzus 2010). The challenge in the para-digm manifests itself in three forms. We will describe these challenges conceptually, beginning with their weakest form and moving on to their strongest form. By weak or strong we do not mean their intellectual strengths but rather how far they deviate from the current paradigm.

The ‘weak form’ of challenge to the paradigm is that the audience to which financial accounting state-ments are addressed are still the capital providers, but these capital providers now require information about an organization’s impact on society and the environment. This is because these impacts can be

financially material. Although the emphasis is on the disclosure of nonfinancial metrics outside of the con-text of accounting statements, the work of SASB is aligned with this challenge.

The ‘semi-strong’ form of challenge to the par-adigm is that the audience of the accounting state-ments are also still the capital providers, but these capital providers now require information independ-ent of financial materiality about an organization’s impact on society and the environment, because cap-ital providers wish to base their decisions not only on risk and return considerations but also consider-ing their impact (Hart and Zingales 2017). Although not focused on accounting statements, the work of impact investment organizations such as the Global Impact Investment Network (GIIN) and the Global Steering Group for impact investing (GSG) are concep-tually aligned with this challenge, as impact investors need impact information to be optimized alongside risk and return.

The ‘strong form’ of challenge is when the ac-counting statements’ audience extends beyond capital providers to include other stakeholders and society at large, and these impacts need to be accounted for as they are directly relevant to those stakeholders. Al-though the emphasis is on the disclosure of ESG-non-financial metrics outside of the context of accounting statements, conceptually the work of GRI is aligned with this challenge. These challenges to the status quo are presented in Table 1.

In turn, the measurement, recognition, and dis-closure of impact-weighted financial accounts, which reflect an organization’s environmental and social impact as well as its financial performance, have re-sulted in the diffusion of a new management prac-

Table 1

Challenges to The Current Accounting Paradigm

Status Quo Weak form of challenge Semi-strong form of challenge

Strong form of challenge

Audience Capital providers Capital providers Capital providers Stakeholders

Audience’s Cares Risk and return Risk and return Risk, return, and impact Risk and return, or risk, return and impact

Needs of Audience Financially material information

Financially material information

Financially and societally material impact

information

Financially and societally material impact

information

Legacy Financial Information

Yes Yes Yes Yes

Financially Material Environmental and Social Impact Information

Yes Yes Yes

Environmental and Social Impact Information Independent of Financial Materiality

Yes Yes

Change From Status Quo

Some environmental and social impact informati-on is financially material

Capital providers include impact investors that

need impact information that is independent of financial materiality

Audience is broader than capital providers

Source: Authors’ illustration.

CONTENT

22 CESifo Forum 3 / 2021 May Volume 22

FOCUS

tice across companies and investors. By 2019, several large organizations around the world, such as Novar-tis, Syngenta, Tata, Safaricom, Roche, Cemento Argos, Lafarge Holcim, and ABN-Amro, had done some form of IWA (Serafeim, Zochowski, and Downing 2019). This diffusion has been accelerated by the creation of the Impact-Weighted Accounts Initiative (IWAI) and com-pany-led member organizations, such as the Value Balancing Alliance (VBA), as well as by accounting firms developing monetary impact measurement methods (PwC 2013).

RESISTANCE TO THE NEW PARADIGM

Kuhn (1962) suggested that all paradigm shifts face strong resistance because the existing paradigm is at risk, and the shift to impact-weighted financial ac-counts is no exception. Skeptics raise two main con-cerns: It cannot be done, and it should not be done.

For historical context, it is important to note that the same criticisms were leveled before the establish-ment of financial accounting standards through the Securities Exchange Act during the Great Depression in the United States. The proposed Generally Accepted Accounting Principles (GAAP) were criticized on the basis that every organization is unique, and that ac-counting is art rather than science. The GAAP were de-rided as imposing substantial costs on companies and thereby threatening growth and capitalism. Of course, hindsight has proven the critics wrong. First, financial accounting standards now exist in every country in the world, guided by International Financial Report-ing Standards (IFRS) in more than 140 jurisdictions and by the US GAAP in the United States. Second, the creation of financial accounting standards was a necessary condition for the creation of large-scale capital markets that have fueled access to finance, growth, and job creation around the world.

Resistance to the weak form used to rest on the argument that investors do not need this information, as it is not financially material. Two developments have settled this argument. The first is evidence that some impacts in some industries can be financially material and price-relevant (Khan, Serafeim, and Yoon 2016; Grewal, Riedl, and Serafeim 2019, Grewal, Hauptmann, and Serafeim 2020; Cheema-Fox et al. 2020). The second is that investors with trillions of dol-lars to invest directly ask for this information because they believe it is financially material (Amel-Zadeh and Serafeim 2018).

The criticism now rests primarily on the argument that impacts cannot be measured. There is merit to this argument, as the measurement of such impacts can be notoriously difficult (Serafeim, Zochowski, and Downing 2019). For example, product impacts can be difficult to measure in some industries (Serafeim and Trinh 2020), and the measurement of well-being effects for employees can be subjective (Freiberg et al. 2020). The response to this criticism is that our ability

to measure impacts with more precision is contingent on taking the first steps to measuring such impacts at all. As in the case of accounting standards, which have experienced several revisions of revenue recognition, or lease liability measurement methodologies, the measurement of impacts will improve over time. This argument, of course, applies across all three forms of the challenge.

The objection to the semi-strong form of the chal-lenge is that investors should not care about impact and that fiduciary duties preclude investors from car-ing about the impact of the companies in which they invest. The validity of this objection is currently being debated among pension funds seeking to understand whether their trustees could or should consider the impact of investments on people and planet. Regard-less of where the legal argument will settle, the reality is that investors have different objectives and that the impact investing market has grown significantly over time. According to GIIN, the impact investing market stood at USD 715 billion in 2019 (GIIN 2020). As the impact investing market grows, the argument that a significant part of the market will require impact information will become stronger.

The objection to the strong form of the chal-lenge is that the audience for accounting statements should be capital providers, not other stakeholders. The problem with this argument is that it rests on normative grounds. From that perspective, the two paradigms can be incommensurable, meaning that it is not possible to understand one paradigm through the conceptual framework and terminology of the rival paradigm (Kuhn 1962). It is likely that the only way to resolve the strong form of the challenge is to first resolve the debate on corporate purpose (Hart and Zingales 2017; Grewal and Serafeim 2020). In other words, if the corporate purpose is to max-imize shareholder value, then one could reject the strong form of the challenge. If corporate purpose includes other stakeholders (Stout 2012; Mayer 2018; British Academy 2019; Henderson 2020; WBCSD 2020), the strong form of the challenge is likely to be a vi-able paradigm guiding the purpose of accounting statements.

Another objection relates to the idea of monetiz-ing impacts in business, a process necessary to have a common denominator and for these impacts to be reflected in financial accounts. The concern is that not everything should be monetized, as doing so might lead to the degradation of morals and human values. Our response is threefold. First, although we concep-tually agree with this objection, there are many im-portant impacts that can be measured without threat-ening our values. For example, impact-weighted ac-counts do not need to reflect the value of human life.1

1 It is relevant to note that regulatory agencies, courts, and policy makers routinely make decisions based on such monetary estimates in areas of healthcare, product safety requirements, and litigation outcomes.

CONTENT

23CESifo Forum 3 / 2021 May Volume 22

FOCUS

Reflecting the monetary impacts of lost productivity and healthcare costs are a possible alternative. There are many impacts that can be reflected in monetary terms, and advances in technology and science allow us to estimate them with improved precision.

The second response relates to effectiveness. The fact that we did not put a price on carbon, our forests, and clean oceans has not resulted in our morals playing a role in taking care of the planet. It has led to a climate catastrophe with a global average atmospheric level of carbon dioxide in 2019 of 409.8 parts per million,2 the highest concentration in 800,000 years; the loss of 35 percent of our man-grove forests, 40 percent of our terrestrial forests, and 50 percent of wetlands (TEEB 2011); also, at least 8 million tons of plastic end up in the oceans every year (Thevenon, Carroll, and Sousa, 2014; Boucher and Friot 2017). The track record of humanity is sim-ilarly abysmal when it comes to other species. Of an estimated eight million animal and plant species, around one million face the threat of extinction (IP-BES 2019), while we have already brought about the extinction of 60 percent of mammals, birds, fish, and reptiles since 1970 (WWF 2018). The lack of a price has done nothing to slow the degradation of the environment, and putting a price on the consump-tion of the natural world could greatly improve the decision-making process. For example, a recent study estimated that each great whale creates cli-mate-related benefits of USD 2 million, representing a value that was previously never considered (Chami et al. 2019).

The third response, which is related to the poor track record just described, is that not measuring these impacts is a value judgement in itself. The value we assigned to them in business has been zero. What we choose to measure reflects our values on what is important and needs to be prioritized. Allowing those impacts to remain invisible creates a perception that they are not important.

VALUE OF DATA AND STANDARDS

In accounting, standards are valuable because they can increase the relevance, reliability, and compara-bility of the reported information. As a result, IFRS and US GAAP are important institutions guiding the production of accounting statements and other im-portant disclosures. While standards have a specific connotation in the field of accounting, we define a standard here as “a methodology that over time be­comes widely accepted and allows data to be linked with other data in a standardized way.”

Under this definition, standards can make data more decision useful and therefore more valuable. Standards empower relationships and transactions,

2 As reported by NOAA (https://www.climate.gov/news-features/understanding-climate/climate-change-atmospheric-carbon-diox-ide).

as they allow people to agree on a set of parameters. For example, the use of Unix time allows people to arrange a time to speak, even if they are living on different continents, while the meter unit allows us to size our real estate, and SAT or GMAT scores en-able an assessment of student performance. There are many standards in the environmental and social space that have increased the value of data. Some examples include measurement methodologies, such as the GHG accounting protocol, which details the methodology and taxonomy for calculating car-bon emissions; certifications, such as the Forest Stewardship Council (FSC), the Rainforest Alliance (RA), or LEED Green Buildings, which detail desirable social and/or environmental practices; and industry initiatives, such as the Extractive Industries Trans-parency Initiative (EITI), which specify disclosures that companies and governments need to make on payments and receipts for the exploitation of natu-ral resources.

A standard creates value because it allows data to link to other data. This makes all the linked data more valuable. Take the example of employment impact. Wages paid to the workforce are one data item. The locations of the employees, the number of employees in each location, and the level of a living wage across different locations are three more data items. Linking the four using the employment impact methodology in Freiberg et al. (2020) makes each of the data items more valuable as it enables an assessment of whether an organization pays above or below the living wage, and what total amount of wages is above or below the living wage.

It is important to realize that perfection is the enemy of standards. The reason is that the value of the standard increases exponentially with adoption, much like in Metcalfe’s Law, in which the effect of a telecommunications network is proportional to the square of the number of connected users of the sys-tem. None of the standards mentioned above — the GHG protocol, FSC, RA, LEED, or EITI — are perfect and all of them have been criticized for different rea-sons. Building a standard that people can easily and widely adopt is more important than trying to create the perfect standard. Easiness and wide adoption are also the reasons why standards need to be open access.

Once the data are connected, using the standard as the basis, they can be analyzed. For example, an analysis of company-level environmental impact data shows that environmental impact has little correla-tion with commercial environmental ratings, but it also reveals that companies with a higher negative environmental impact have a lower market valuation in industries such as building products, independent power producers, industrial conglomerates, metals and mining, and chemicals (Freiberg et al. 2020). The analysis also finds that environmental impact has become more financially material over time. From

CONTENT

24 CESifo Forum 3 / 2021 May Volume 22

FOCUS

a data science perspective, the next steps would be to form predictions with the ultimate objective of assisting with decisions.

THE FUTURE

As our ability to account for environmental and social impacts increases, so will the ability of managers and investors to make better decisions and the ability of policy makers to design better policies. Our ability to make progress requires us to solve several unre-solved issues.

First, there must be a regulatory organization that defines the scope of the environmental and so-cial impacts that need to be accounted for to miti-gate scope bias. The European Commission is already taking steps to define scope in sustainable finance through the development of a technical expert group that seeks to create a taxonomy for the environmen-tal impact of economic activities and a green bond standard. Importantly, missing vital impacts might positively or negatively bias the impacts that differ-ent organizations might have, depending on what is included. For example, when evaluating the impact of consumer-packaged goods companies on consumers, the exclusion of the positive health effects of calcium consumption can give rise to more negative impacts (Serafeim and Trinh 2021).

The scope of accounted impacts will define what is prioritized and valued by companies and their in-vestors. Of course, such scoping of impacts is al-ready common practice in the ESG investing space. It is important to recognize that any evaluation that aggregates ESG metrics using weights to arrive at an overall assessment de facto prioritizes some issues over others, thereby making judgements about their importance.

Second, different impacts have different meas-urement bases, such as impacts on employees rela-tive to impacts on the environment. This can give rise to incomparable numbers across different impacts. While employees could be evaluated using wages as the measurement base, the environment might be evaluated using estimates of damage on human health and abiotic resources. Of course, this problem is not new. The financial accounting system has long faced similar problems, such as determining when to ac-count for items at historical cost and when to do so at fair market value.

Third, an accounting treatment is needed that creates both stock and flow measures, as both could be important for accountability and contract-ing purposes. The former becomes part of an im-pact-weighted balance sheet while the latter becomes part of an impact-weighted income statement. Neg-ative impacts can accumulate in a negative equity reserve or a liability account and positive impacts in a positive equity reserve or an asset account. For ex-ample, environmental impacts could be accumulated

in an environmental liability number while positive employment impacts on a labor investments asset.

THE OPPORTUNITY

The first decade of the 21st century showed that un-measured and unmanaged financial risks can bring the world to its knees. The second decade gave rise to a response that, due to the rapidly deteriorating en-vironmental and social conditions, we need to build a more sustainable economy and inclusive society. The third decade represents an opportunity to create the guardrails and accountability mechanisms to make this a reality. Transparent, scalable, and comparable measurement of environmental and social impacts will be a necessary condition. In turn, the incorpora-tion of these measures into outcome-based financing, compensation, and policies could well turn out to be a condition that is sufficient to alter our behavior in an unprecedented way.

REFERENCES Amel-Zadeh, A. and G. Serafeim (2018), “Why and How Investors Use ESG Information: Evidence from a Global Survey”, Financial Analysts Journal 74(3), 87–103.

British Academy (2019), Principles for Purposeful Business – How to Deliver the Framework for the Future of the Corporation. An Agenda for Business in the 2020s and Beyond.

Bullard, N. (2021), The Sustainable Debt Market is All Grown Up, Bloomberg Green, January 14, https://www.bloomberg.com/news/arti-cles/2021-01-14/the-sustainable-debt-market-is-all-grown-up (accessed March, 2021).

Chami, R., T. Cosimano, C. Fullenkamp, S. and Oztsun (2019), “Nature’s Solution to Climate Change”, Finance & Development 56(4), 34–38.

Cheema-Fox, A., B.R. LaPerla, G. Serafeim, D. Turkington, and H.S. Wang (forthcoming), “Decarbonizing Everything”, Financial Analysts Journal.

Christensen, D., G. Serafeim, and A. Sikochi (forthcoming), “Why is Cor-porate Virtue in the Eye of the Beholder? The Case of ESG Ratings”, The Accounting Review.

Eccles, R.G. and M.P. Krzus (2010), One Report: Integrated Reporting for a Sustainable Strategy, John Wiley & Sons.

Flammer, C., B. Hong, and D. Minor (2019), “Corporate Governance and the Rise of Integrating Corporate Social Responsibility Criteria in Execu-tive Compensation: Effectiveness and Implications for Firm Outcomes”, Strategic Management Journal 40(7), 1097–1122.

Freiberg, D., K. Panella, G. Serafeim, and T. R. Zochowski (2020), “Accounting for Organizational Employment Impact”, Harvard Business School Impact-Weighted Accounts Research Report.

GIIN (2020), Annual Impact Investor Survey, Report.

Grewal, J., E. Riedl, and G. Serafeim (2019), “Market Reaction to Manda-tory Nonfinancial Disclosure”, Management Science 65(7), 3061–3084.

Grewal, J. and G. Serafeim (2020), “Research on Corporate Sustainabil-ity: Review and Directions for Future Research”, Foundations and Trends® in Accounting 14(2), 73–127.

Grewal, J., C. Hauptmann, and G. Serafeim (forthcoming), “Material Sustainability Information and Stock Price Informativeness”, Journal of Business Ethics.

Hart, O. and L. Zingales (2017), “Companies Should Maximize Share-holder Welfare Not Market Value”, Journal of Law, Finance, and Account­ing 2, 246–274.

Henderson, R. (forthcoming), “Innovation in the 21st Century: Architec-tural Change, Purpose, and the Challenges of Our Time”, Management Science.

IPBES (2019), Summary for Policymakers of the Global Assessment Re-port on Biodiversity and Ecosystem Services of the Intergovernmental Science–Policy Platform on Biodiversity and Ecosystem Services, Report.

Khan, M., G. Serafeim, and A. Yoon (2016), “Corporate Sustainability: First Evidence on Materiality”, Accounting Review 91(6).

CONTENT

25CESifo Forum 3 / 2021 May Volume 22

FOCUS

Kuhn, T. (1962), The Structure of Scientific Revolutions, University of Chicago Press.

Macve, R.H. (2015), “Fair Value vs Conservatism? Aspects of the History of Accounting, Auditing, Business and Finance from Ancient Mesopota-mia to Modern China”, British Accounting Review 47(2), 124–141.

Mayer, C. (2018), Prosperity: Better Business Makes the Greater Good, Oxford University Press.

PwC (2013), Measuring and Managing Total Impact: A New Language for Business Decisions, Report.

Stout, L.A (2012), The Shareholder Value Myth: How Putting Sharehold­ers First Harms Investors, Corporations, and the Public, Berrett-Koehler Publishers.

Serafeim, G., T. R. Zochowski, and J. Downing (2019), “Impact-Weighted Financial Accounts: The Missing Piece for an Impact Economy”, White Paper, Harvard Business School, Boston, MA.

Serafeim, G. and K. Trinh (2020), “A Framework for Product Im-pact-Weighted Accounts”, Harvard Business School Impact-Weighted Accounts Research Report.

Soll, J. (2014), The Reckoning: Financial Accountability and the Rise and Fall of Nations, Basic Books (AZ).

TEEB (2011), The Economics of Ecosystems and Biodiversity in National and International Policy Making, edited by Patrick ten Brink, London and Washington: Earthscan.

WBCSD (2020), Reinventing Capitalism: a Transformational Agenda, Vision 2050 Issue Brief.

WWF (2018), Living Planet Report 2018: Aiming Higher, edited by M. Grooten and R.E.A. Almond, Gland.

CONTENT

26 CESifo Forum 3 / 2021 May Volume 22

FOCUS

Hao Liang and Jasper Teo

Peer Effects of Commitment to ESG Goals – How Stakeholders Affect One Another in the Ecosystem

Companies today are becoming more socially re-sponsible and have begun to integrate corporate social responsibility (CSR) into their business mod-els. CSR refers to a firm’s commitment to pursue business objectives that lead to positive social and environmental outcomes rather than focusing solely on maximizing economic goals. It has become in-creasingly important for companies to commit to CSR goals to ensure that their business remains eth-ical and sustainable in the long term. Research has shown that CSR has been able to reduce risk and increase customer satis faction and market value. CSR strategies carried out by stakeholders in an eco-system can have a significant and material impact on peer firms. Peer firms are pressured to commit to higher CSR standards set by their clients, suppli-ers or competitors. This article reviews three major channels through which peer effects of corporate commitment to ESG goals, as documented in the academic literature (some of which are written by the author of this article), can take place: product market competition, global supply chain, and insti-tutional investors’ propagation.

PEER EFFECTS ON COMPETITORS: CSR STRATEGIES WILL HAVE A MATERIAL IMPACT ON THE STOCK PRICE OF COMPETITORS

CSR policies have become widely adopted by compa-nies across the world. In the United States, the Gov-ernance & Accountability Institute (G&A) found that the number of S&P500 companies that release sus-tainability reports reached an all-time high of 90 per-cent in 2020, compared to a figure of only 20 percent

in 2011. Despite the widespread publishing of CSR re ports, there are still doubts as to whether this in-crease is due to investors and companies reacting to peers’ CSR strategies or whether there are other fac-tors at play which influence the company’s decision to adopt CSR initiatives.

The underlying argument is that companies do not operate in silos. Their actions are often observed and replicated by peers. If CSR can indeed increase firm value, a company’s adoption of CSR will put it at a competitive advantage and its competitors at a competitive disadvantage. When a company in-troduces CSR strategies, peers may see it as a threat and respond strategically by adopting new CSR ini-tiatives so as not to be left behind. An example is a firm which invested in an environmentally friendly technology that is able to reduce carbon emission in its production processes. By utilizing an environ-mentally friendly technology, a firm is able to hu-manize its brand and develop brand empathy with socially conscious consumers. Yet, despite the numer-ous advantages of doing so, companies may not want to shift to an environmentally friendly technology, as this can potentially lead to margins being eroded from their bottom line. However, when one company decides to challenge the status quo and invest in the technology, it will be able to gain a competitive ad-vantage over its peers. As a result, peers who believe that they will lose their edge may invest in this en-vironmentally friendly technology to ensure that it remains competitive in the industry.

Empirical Evidence

The authors of a paper titled “Peer Effects of Corporate Social Responsibility” (Cao, Liang, and Zhan 2019) investigated how listed com-

panies responded to competitors’ adoption of CSR-related shareholder proposals that are being voted during shareholder meet-

ings. The study tested more than 3,000 US public nonvoting peer firms in the period be-tween 1997 and 2011. They compared the ef-fects of a firm’s shareholder-sponsored CSR proposals at annual meetings that either pass or fail by a small margin (around the 50 percent threshold) with the product-market peer firms’ response in terms of ESG com- mitment. This is to reproduce a randomized assignment of CSR proposals to companies and, hence, ensure that it is not correlated

is Associate Professor of Finance at Singapore Management Uni-versity, PGR Coordinator, Fi-nance, DBS Sustainability Fellow. His research interests include corporate finance, corporate gov-ernance, corporate social respon-sibility, and law and finance, and international business.

is Student Assistant at Singapore Management University.

Hao Liang Jasper Teo

CONTENT

27CESifo Forum 3 / 2021 May Volume 22

FOCUS

with peer-firm characteristics. The underlying idea is that there is no reason to expect that a peer firm of a company that marginally passes a CSR proposal is systematically different from a company in which a CSR proposal marginally fails. Hence, these ‘close-call’ CSR proposals provide a good proxy of a random variation of a firm’s commitment to ESG that can be used to estimate causal ESG peer effects.

The authors observed two key results. (1) The passage of a close-call CSR proposal by a firm led to the adoption of a similar CSR practice by its com-petitors the following year, as measured by their corresponding CSR scores. (2) Firms that passed a close-call CSR proposal were able to achieve a higher three-day cumulative abnormal return (CAR) and sub-sequently a higher CSR score, whereas its competitors experienced a lower CAR around the same dates. This indicates that firms that were unable to catch up with peers that pass a close-call CSR proposal suffered a significantly lower stock return, which is consistent with the notion that CSR is a strategic move in re-sponse to a threat by other firms. Given the possibility that some of the proposals were not implemented, the authors conducted a textual analysis to disentangle the effect of passage and actual implementation. They found that the effects were even stronger if the close-call CSR proposal was implemented in the following year (based on news reports), which suggests that the effects were not simply a signal but could in fact have a material impact on peer firms.

PEER EFFECTS ON SUPPLIERS: CUSTOMERS ARE ABLE TO EXERT CONSIDERABLE INFLUENCE ON THE CSR STANDARDS OF THEIR SUPPLIERS

Another mechanism through which ESG commit-ments can spill over among peers is the supply chain. Anecdotal evidence has suggested that as corpo rations develop their CSR policies, corporate customers become increasingly concerned not just with their own CSR standards, but also those of their suppliers. As more companies shift their supply chain overseas to gain a competitive advantage, corpo-rations finding it challenging to enforce their CSR standards are on their supplier networks. Despite the high cost of ensuring that their global supply chain meets the CSR standards, there are several reasons why corporations are driven to do this. Motivations for pushing suppliers to employ CSR strategies in-clude appeasing stakeholders, avoiding negative publicity, retaining employees and attracting new customers who are interested in purchasing products from firms that themselves purchase from sustain-able sources.

Empirical Evidence

Using the Factset Revere database of firm-level net-works of customers and suppliers around the world

and Thomson Reuters’s ASSET4 database of firm-level ESG ratings, the authors of a recent study ti-tled “Socially Responsible Corporate Customers” (Dai, Liang, and Ng 2020) obtained a sample comprising 34,117 unique corporate customer-supplier pairs from 50 countries worldwide for the period 2003 to 2015. They found evidence that customers made an effort to ensure that suppliers met certain CSR standards while not exhibiting any influence on customers’ CSR standards.

However, a potential issue is that even though there are strong correlations between customer CSR and subsequent improvements in suppliers’ CSR practices, it may be because customers tend to se-lect suppliers that are more likely to cooperate and commit to high CSR standards. To circumvent this issue, the authors examined the effect of customers on suppliers’ CSR standards by employing a regres-sion discontinuity approach that relies similarly on the passage of close-call proposals. The results sug-gest that the passage of a customer’s close-call CSR proposal will indeed lead to their suppliers adopting similar CSR initiatives, as evidenced by the increase in the supplier’s CSR score in the following year, as compared to a supplier for whom the customer’s CSR proposal fails by a small margin. In another test, the authors looked at companies that were affected by unexpected product safety scandals, such as the 2008 Chinese milk scandal and the 2013 Toyota car/Takata airbag recalls. The results showed a positive correlation between a customer’s product respons-ibility rating and that of its supplier, which became even stronger for firms in the relevant industry and on the relevant markets, after the year of the scan-dal. This suggests that the improvement in CSR could be attributed to customers pushing their suppliers to improve their behavior in response to the scandal.

PEER EFFECTS OF INSTITUTIONAL OWNERSHIP:SHAREHOLDERS HAVE AN IMPORTANT ROLE TO PLAY IN DRIVING ESG CHANGE WITHIN A FIRM

Institutional investors play a huge role not only in the practice of responsible investing, but also in the propagation of ESG commitments among their portfolio companies through both engagement and voting by feet. There has been increasing pressure by capital owners on institutional investors to en-sure that they consider ESG issues when deciding how to deploy their capital. Investors today assess companies not only on the potential financial gains to be had from their investments, but also on the environmental and social (E&S) outcome that can be achieved.

Empirical Evidence

Various studies have been conducted on respon-sible investing. The authors of a working paper ti-

CONTENT

28 CESifo Forum 3 / 2021 May Volume 22

FOCUS

tled “Responsible Institutional Investing Around the World” (Gibson, Glossner, Krueger, Matos, and Ste-ffen 2020) used survey data from the UN Principles for Responsible Investment (UN PRI). UN PRI is the world’s largest group of institutional investors that encourages signatories to engage in responsible in-vestment to enhance returns and manage risks more effectively. The authors of another study, titled “Do Institutional Investors Drive Corporate Social Responsi­bility?” (Dyck, Lins, Roth, and Wagner 2019) obtained data from the Thomson Reuters ASSET4 database for firm-level ESG ratings and the FactSet database for firm-level institutional ownership. The samples comprise 19,849 firm-year observations and cover 3,277 firms from 41 countries in the period 2004–2013. Three interesting observations can be made in these two studies with regard to how institutional investors propagate ESG practices among its portfo-lio companies.

The first observation is that institutional inves-tors tend to propagate better ESG practices by being signatories of the UN PRI. This is likely due to the signatories’ desire to demonstrate their commitment to achieving the six ‘Principles of Responsible In-vestment’. These include incorporating ESG issues into the investment analysis, engaging with firms to improve ESG practices, and seeking disclosures from portfolio companies on ESG-related issues. As such, it is likely that signatories will take greater re-sponsibility and actively manage ESG issues in their portfolio. Research has shown that the ESG impact of institutional investors who are signatories of the UN PRI is greater than those who did not sign. How-ever, UN PRI signatories exhibit different levels of ESG practice depending on the country in which they are based. UN PRI members in countries out-side the United States exhibit better ESG footprints than investors based in the United States. It is even more surprising that signatories based in the United States tend to exhibit ESG footprints that are no bet-ter than those of non-signatories. Despite United States based investors forming the largest group of new PRI signatories in recent years, there are no evidence which show that they improve their ESG footprints after signing the PRI. It was also observed that if they only implement a partial ESG strategy (applying it only to a fraction of their total assets under management), these signatories exhibit even worse ESG footprints than their non-signatory coun-terparts. Furthermore, the authors found that United States-based investment managers mainly serve the retail segment rather than institutional clients who monitor their managers more closely. This suggests that some United States based signatories are engag-ing in ‘greenwashing’ to attract more capital from ESG-conscious investors.

The second observation is that institutional in-vestors propagate better ESG practices by using se-lection criteria and a strong voice. The first mech-

anism is the threat of exit or of investing only in firms with strong E&S policies. The second is to use their shareholding voice to engage with the man-agement. Institutional investors are able to utilize several methods to determine whether to invest or to exit an investment, which include using negative screening to exclude poor E&S-performing firms or positive screening to invest only in firms with cer-tain E&S standards. However, this does not always lead to better returns. In the first study, the authors found evidence that certain selection methods such as negative screening, integration, and engagement reduces portfolio risk but does not enhance returns, and they concluded that ESG strategies mainly serve risk-management purposes. Furthermore, the insti-tutional investors’ voice is a powerful mechanism in pushing for E&S change. Having a dominant voice is important when being part of an investor organiza-tion such as the UN PRI, which advocates active en-gagement of its signatories with firms in order to en-hance their E&S performance. More importantly, the authors found that successful proposals are typically not voted on and suggest that E&S shareholder pro-posals are primarily driven by private negotiations. The authors conclude that private engagement is the main channel through which investors push firms to enhance their E&S standards.

The third observation is that institutional in-vestors propagate ESG practices by driving firms to lower risk. Institutional investors drive firms to im-prove their E&S performance so as to achieve better financial returns. Institutional investors that focus on E&S activism are concerned with managing long-term risk and aim at achieving a long-term return on their investment. E&S investment is able to provide insurance against event risk and product market dif-ferentiation. Furthermore, Evidence that investors are motivated by financial returns can be seen in the Deepwater Horizon crisis, in which the institutional investors’ push for better E&S performance only im-pacted the environmental aspect and only for extrac-tive industries. The goal of institutional investors in this case was to reduce the potential of another in-cident occurring in the future that would cause them to incur unexpected costs.

CONCLUSION

This article looks at the key channels through which peer effects of commitment to ESG goals can take place. The channels and impacts mentioned in this article are by no means exhaustive, but the benefits of commitment to ESG goals are immense. Commit-ment to ESG goals has been shown to create value, as it lowers risk and generates positive externalities to the environment. An important policy implica-tion of these findings is that there is a ripple effect from improving one company’s ESG commitment, as it will likely propagate to other companies through

CONTENT

29CESifo Forum 3 / 2021 May Volume 22

FOCUS

product market competition, the global supply chain, and institutional shareholding. Such ESG spillover can trigger a multiplier effect. From a managerial perspective, it is important that a firm takes into account peer firms’ ESG policies when designing its own strategy and CSR policy. However, despite the vast improvement in stakeholder commitment to ESG goals, there are still more measures that could be taken by firms to influence peers in improving their ESG footprints. These may include firms setting a high CSR standard as an industry benchmark, cus-tomers placing social pressures on their suppliers to commit to ESG goals, and investors directing their

fund managers to implement a complete ESG strategy instead of engaging in ‘greenwashing’.

REFERENCES Cao, J., H. Liang, and X. Zhan (2019), “Peer Effects of Corporate Social Responsibility”, Management Science 65(12), 5487–5503.

Dai, R., H. Liang, and L. Ng (forthcoming), “Socially Responsible Corpo-rate Customers”, Journal of Financial Economics.

Dyck, A., K. V. Lins, L. Roth, and H. F. Wagner (2019), “Do Institutional Investors Drive Corporate Social Responsibility? International Evidence”, Journal of Financial Economics, 131(3), 693–714.

Gibson, R., S. Glossner, P. Krueger, P. Matos, and T. Steffen (2020), “Re-sponsible Institutional Investing Around the World”, Swiss Finance Insti-tute Research Paper No. 20-13, European Corporate Governance Insti-tute – Finance Working Paper 712/2020.

CONTENT

30 CESifo Forum 3 / 2021 May Volume 22

FOCUS

Fighting climate change and promoting sustainable development are currently among the most impor-tant social, economic and political concerns. At the European level, these issues have long since come to the fore, not least in the European Green Deal. By presenting the EU Action Plan on Sustainable Fi-nance, the European Commission has also assigned a key role to the financial system. The aim of the Action Plan is to regulate the financial system in such a way that it is able to contribute to sustain-able development. All member states are required to implement the EU Action Plan; national govern-ments may be even more ambitious. The German government, for instance, has set itself the goal of establishing Germany as a leading center for sus-tainable finance. It was against this background that the Federal Government’s Sustainable Finance Committee was set up. The German Federal Finan-cial Supervisory Authority (BaFin) has already begun to integrate the issue of sustainability into German banking supervision.

While the need for a more sustainable economy is hardly disputed, opinions differ on the question of what measures should be taken. Numerous measures have been discussed in the EU and on the national level – some of them controversial. As in many Euro-pean countries small and medium-sized enterprises (SMEs) are of particular importance, the impact which the new regulations have on the financial sector and ultimately on SME financing is crucial, albeit hitherto little discussed.

THE EU ACTION PLAN AND MEASURES AT EUROPEAN LEVEL

The objectives of the EU Action Plan can be put into three categories (European Commission 2018):

1. Directing capital flows into sustainable invest- ments;

2. Managing financial risks arising from environmen-tal and social issues;

3. Promoting transparency and a long-term view in financial and economic activity.

The EU Action Plan comprises a catalog of ten meas-ures; its aim is to anchor sustainability in the financial sector. The taxonomy forms a central instrument for assessing the sustainability of an economic activity and can be considered the core of the Action Plan. It creates a uniform definition of sustainability as a basis for assessing green financial products. The taxonomy classifies certain economic activities as sustainable. The aim is to lay down criteria for green financial products, to ensure the necessary market transparency for investors, and to avoid the problem of ‘greenwashing’.

In addition to developing the taxonomy, the Ac-tion Plan introduces an EU label for green financial products, establishes sustainability obligations for asset managers and institutional investors, strength-ens the transparency of firms regarding their environ-mental, social, and governance (ESG) policies, and considers adding a ‘green supporting factor’ in EU prudential rules for banks and insurance companies (European Commission 2020a).

For economic activities to be considered envi-ronmentally sustainable, they must not only make a significant contribution to at least one of the six

environmental goals, but they

Christa Hainz, Johann Wackerbauer and Tanja Stitteneder

Economic Policy Goals of the Sustainable Finance Approach: Challenges for SMEs*

* This article is a summary of the study “Sustainable Finance – Eine kritische Würdigung der deutschen and europäischen Vorhaben,” conducted by the ifo Institute on behalf of the Chamber of Industry and Commerce for Munich and Upper Bavaria (Fuest, Hainz, Wacker-bauer, and Stitteneder 2020), and an updated translation of Hainz, Stitteneder, and Wackerbauer (2020).

is Deputy Director of the ifo Center for International In-stitutional Comparisons and Migration Research, focusing on institutions and financial intermediation.

is Senior Researcher of the ifo Center for Energy, Climate and Resources. His main areas of expertise are environmental eco-nomics, environmental policy and eco-industry.

is Project Specialist at the ifo Center for International Institu-tional Comparisons and Migra-tion Research. Her work focuses mainly on EU policy and interna-tional relations.

Christa Hainz Johann Wackerbauer Tanja Stitteneder

CONTENT

31CESifo Forum 3 / 2021 May Volume 22

FOCUS

must not harm the other environmental goals (‘Do no significant harm’ – DNSH). They also have to ensure a minimum level of protection for employees (TEG 2019a).1 The taxonomy defines technical screening criteria for the first two items in particular. Ideally, these thresholds must be met for an activity to be considered sustainable.

OBJECTIVES AND DISCUSSION OF MEASURES

In the following we discuss how the proposed meas-ures can contribute to achieving the objectives pre-sented above.

Goal 1: Channeling Capital Flows into Sustainable Investments

Substantial investments are needed to finance the transition to greater sustainability. In the EU, it is es-timated that EUR 260 billion of additional investment will be needed annually for climate action alone (Euro-pean Commission 2021). Investing in sustainable eco-nomic activities avoids creating new assets that are, for instance, exposed to transition risks and devalued by political decisions to combat climate change.2 It is estimated that currently over 40 percent of pension fund and mutual fund equity portfolios could be af-fected by transition risks (Monasterolo 2020).

To steer capital flows more toward sustainable investments, sustainability is being specified as a cri-terion in investment advice and lending. This leads to new distinctions being made in financing options and conditions. This could mean that sustainable firms are better able to finance themselves than less sus-tainable firms. In relative terms, therefore, the cost of capital increases for less sustainable firms and firms that do not (or cannot) adequately demonstrate their sustainability, and their profitability decreases. This is reflected in their investment and production deci-sions. This is one way in which the criterion of sus-tainability has an impact on the real economy via the financial market.

Previous empirical studies have looked at invest-ment performance and credit conditions. For example, a meta-analysis that merges the results of 2,000 stud-ies concludes that the vast majority of studies find a higher return on sustainable investments. This is par-ticularly true for investments in sustainable firms, but less so for investments in portfolios, such as mutual funds (Friede, Busch, and Bassen 2015). One might argue that investors do not require any additional incentives to put their capital into sustainable invest-ments if they are more profitable. A recent study has,

1 The six environmental goals are: climate change mitigation, cli-mate change adaptation, sustainable use and protection of water and marine resources, transition to a circular economy, waste pre-vention and recycling, pollution prevention and reduction, and pro-tection and restoration of biodiversity and ecosystems (TEG 2019a).2 Transition risks are risks arising from the transition to CO2-free economic structures.

however, shown that retail investors, in contrast to institutional investors, reduced their investments in sustainable mutual funds after the outbreak of the COVID-19 pandemic (Dottling and Kim 2021).

Investigations into investment performance and lending are based on sustainability ratings, which firms opt for voluntarily. The performance of sustain-able investments can be better for several reasons. One reason is that ESG scores capture firm charac-teristics that were not previously transparent and that (help) explain a firm’s profitability. It could also be that more profitable firms opt for a sustainability rating anyway. This selection effect would not exist if all firms were required to provide a sustainabil-ity rating. In addition, financial market participants could increase the profitability of firms through their investment by exerting pressure on the management and (potentially) withdrawing their capital.

Furthermore, sustainable investments could have a positive externality on other investments if - through international coordination - the transition to a more sustainable real economy can be realized regionally at lower cost (Kittner et al. 2017). The un-derlying mechanism results from the learning curves observed in real terms for photovoltaics, wind energy, and battery storage – the key technologies of the en-ergy transition. This externality should be addressed in a targeted and appropriate way, for example, by granting subsidies for investments and not for their financing.

Goal 2: Managing Environmental Risks

Involving the financial sector in environmental pol-icy, and climate policy in particular, is relatively new. It is therefore important to clarify whether financial market products can be effective at all as a com-plement to a classic environmental or climate pro-tection policy. The theoretical foundation of envi-ronmental and climate protection policy is that any environmental damage caused or to be avoided must be taken into account in the polluter’s cost calcula-tion. Therefore, the environmental policies of the Eu-ropean Union and the Federal Republic of Germany are essentially based on the ‘polluter pays’ principle. The consideration of external environmental costs changes the cost structure to the disadvantage of environmentally harmful production and consump-tion methods and in favor of more environmentally friendly alternatives. In the case of price-elastic de-mand functions, this results in a decrease in demand for the environmentally harmful alternative and an increase in demand for the more environmentally friendly alternative. Production is adjusted and, in the longer term, investment is redirected. In other words, the real economic adjustment to the use of environmental policy instruments already achieves the desired steering effect, and the financial adjust-ments in the corporate sector follow the real eco-

CONTENT

32 CESifo Forum 3 / 2021 May Volume 22

FOCUS

nomic ones. This raises the question of whether the financial sector has any role to play at all in environ-mental and climate protection policy.

One possibility is that sustainable financial in-vestments serve as a support for traditional envi-ronmental policy. Such support may be necessary if the effect of the environmental relief brought about by the policy instruments is in the desired direction but the envisaged environmental target, for exam-ple the reduction target for greenhouse gases, is not fully achieved. In this case, intervention in the finan-cial sector with criteria for ‘green financial products’ could, in theory, be supportive. It is unclear what contribution this would make to the environmental objectives, as the level of demand for green finan-cial products would remain uncertain. The reduction target would be certainly achieved if all emitters of greenhouse gases were included in the EU emissions trading system. However, in this case it would no longer be necessary to support the effect through financial market regulation.

Goal 3: Promoting Transparency and a Long-Term View in Economic and Financial Activity

This objective of the EU Action Plan addresses a cen-tral problem of corporate governance. If ownership and control of a firm are not in the same hand, the different levels of information among the parties in-volved (asymmetric information) can create incentive problems. This means that the manager’s goals do not always coincide with those of the owner. Incentive problems can arise, for example, when a manager’s plans have a shorter time horizon than the owner’s. The EU Action Plan addresses these two problems. On the one hand, information asymmetries are re-duced through greater transparency. On the other hand, prolonging the decision-making time horizon aims at promoting the long-term nature of economic activity. It is important to note that incentive prob-lems play a much greater role for large firms financed via the capital market than for smaller firms that are owner- or family-run.

The Non-Financial Reporting Directive (NFRD) ex-tends the reporting obligations of firms to non-finan-cial aspects to strengthen the transparency of firms in terms of their sustainability. Since 2018, larger firms have to report on such topics as environmental pro-tection, social responsibility and governance. Going forward, the reporting obligations will be extended to include climate-related information, following the recommendations of the Task Force on Climate-Re-lated Financial Disclosures (TCFD) established by the G20 Financial Stability Council (European Commission 2020b). The taxonomy also requires these firms to re-port on the sustainability of their economic activities (European Commission 2019).

A broad information base creates an awareness of the problem in firms and enables them to bet-

ter manage their sustainability issues. At the same time, there are data collection costs that firms will only be prepared to incur if the resulting benefits are higher than the costs – unless they are obliged to do so. For investors, increased transparency cer-tainly helps when it comes to assessing the sustain-ability of a potential investment. The extent to which transparency promotes long-term decision-making depends on the ability of investors to expand a man-ager’s decision-making horizon. If the investor has a longer decision-making horizon than the manager, it is important how much influence the investor has on decision-making in the firm. Here too, more infor- mation should prove to be helpful. Ultimately, it is the actually available corporate governance mech-anisms and, of course, the financing structure of the firm that will determine how much influence in-vestors can actually exert. However, if it is uncertain how sustainability criteria will develop and whether the firm can fulfil these criteria in the long term, capital providers, especially banks, will also be af-fected. One possible reaction could be that they shorten their investment horizon, which could, for example, lead to offering loans with shorter maturities.

Politicians can also play a role in longer term decision-making. By establishing reliable framework conditions and regulatory requirements that can be foreseen in the long term, economic policy can re-duce the level of uncertainty for firms and thus help with expectation management. For firms, this reduces the risk of investing in assets that might lose much of their value as a result of political decisions. Re-ducing uncertainty creates incentives for sustainable investment.

CHALLENGES FOR SMES

The sustainable finance approach poses challenges for SMEs in two respects. First, the new regulatory requirements must be in reasonable proportion to the firm’s size. Second, they must take into account the special nature of the main bank relationship that many SMEs have.

Proportionality of Requirements and Firm Size

According to the EU Action Plan, sustainability dis-closure is voluntary for firms. This means that SMEs are able to opt out – especially since financing via the capital market is of less relevance to them. However, firms that are subject to the Non-Financial Report-ing Directive, i.e., public interest entities with more than 500 employees, are obliged to disclose the sus-tainability of their economic activity according to the taxonomy. However, from the taxonomy the extent of this obligation and whether it also applies to the supply chain is unclear. If SMEs that are suppliers of firms subject to the disclosure requirement were also

CONTENT

33CESifo Forum 3 / 2021 May Volume 22

FOCUS

obliged to report, they would be disproportionately affected by these costs, as they do not increase in proportion to the firm’s size but include a consider-able fixed component.

Since SMEs act as both suppliers and buyers within supply chains, there are multiple points at which they would be affected by this requirement. Even the taxonomy has identified proportionality as an issue (TEG 2019b). Policymakers could support SMEs by applying the disclosure requirement only above a certain firm size or by compensating firms financially for the effort they are required to expend. In addition, all the options for making relevant infor-mation publicly available should be exhausted, for example, concerning emission levels, to minimize the costs of non-financial reporting.

The Special Nature of the Main Bank Relationship

According to the Non-Financial Reporting Directive, major banks but also the central institutions of sav-ings banks and of cooperative banks must disclose the sustainability of their investment and loan portfolios. This means that banks have to gather this information for individual loans (BaFin 2019). In this context, it should be noted that the main bank relationship is of great importance to many firms and to SMEs in particular. A survey for Germany shows that, on av-erage, a firm has maintained its relationship with its most important bank for 29 years (Hainz and Wiegand 2013). Both parties build a reputation as a reliable ne-gotiating partner through this main bank relationship, which can be considered as an implicit contract. This implicit contract creates an incentive for firms to dis-close information that is relevant to the bank’s credit assessment. With respect to the new regulations the question is to what extent and in what form the bank is required to record information on sustainability risks and what effort this entails for both the bank and the borrower.

POLICY RECOMMENDATIONS

With climate change, the goal of a more sustaina-ble economy is increasingly the focus of the public debate. The EU Action Plan on Sustainable Finance attributes a crucial role to the financial system. It is widely accepted that a sustainable economy requires sustainable investments. However, how this can be implemented and what exactly is meant by sustaina-bility is currently heavily debated. Since SME financing plays a particularly important role in many European countries, it is all the more important to integrate the perspective of SMEs into the design of the EU Action Plan. This is specifically expressed in the following recommendations:

‒ Policymakers should create incentives for the transition to a more sustainable economic

system, by establishing reliable framework conditions.

‒ As intended, the application of the taxonomy to financial assets should be neither mandatory nor applied to all financial products.

‒ Proportionality for SMEs is crucial. Additional in-formation requirements place a disproportionate burden on SMEs. Therefore, SMEs should be ex-empted from reporting and disclosure require-ments or relieved of the additional costs.

‒ Information should be allowed to flow within the main bank relationship without any burdensome documentation requirements, so that the full ad-vantage of the long-term main bank relationship can unfold.

REFERENCES BaFin (2019), Merkblatt zum Umgang mit Nachhaltigkeitsrisiken, Bun-des anstalt für Finanzdienstleistungsaufsicht, Bonn.

Dottling, R. and S. Kim (2021), Sustainability Preferences Under Stress: Evidence from Mutual Fund Flows During COVID-19, Erasmus Platform for Sustainable Value Creation.

European Commission (2018), Frequently Asked Questions: Commission Proposals on Financing Sustainable Growth, Brussels, https://ec.europa.eu/commission/presscorner/detail/en/MEMO_18_3730, May 24, 2018.

European Commission (2019), Questions and Answers: Political Agree-ment on a System for Sustainable Investments (Taxonomy), Brussels, December 18, 2019.

European Commission (2020a), Sustainable Financing, https://ec.europa.eu/info/business-economy-euro/banking-and-finance/green-finance_de (accessed on February 14, 2020).

European Commission (2020b), Non-Financial Reporting, https://ec.europa.eu/info/business-economy-euro/firm-reporting-and-au-diting/firm-reporting/non-financial-reporting_de (accessed on Febru-ary 14, 2020).

European Commission (2021), Overview of Sustainable Finance, https://ec.europa.eu/info/business-economy-euro/banking-and-finance/sustainable-finance/overview-sustainable-finance_de (accessed on April 1, 2021).

Friede, G., T. Busch, and A. Bassen (2015), “ESG and Financial Perfor-mance: Aggregated Evi-dence from More than 2000 Empirical Studies”, Journal of Sustainable Finance & Investment 5(4), 210–233.

Fuest, C., C. Hainz, J. Wackerbauer, and T. Stitteneder (2020), Sustain-able Finance – Eine kritische Würdigung der deutschen und europäis-chen Voraben, study by the ifo Institute on behalf of the Chamber of Industry and Commerce for Munich and Upper Bavaria.

Hainz, C., J. Wackerbauer, and T. Stitteneder (2020), “Der Sustainable- Finance-Ansatz: Entwicklung, wirtschaftspolitische Ziele und besondere Herausforderungen für den Mittelstand”, ifo Schnelldienst 73(10), 10–14.

Hainz, C. and M. Wiegand (2013), “Financing the German Economy Dur-ing the Financial Crisis”, CESifo DICE Report 11(1), 48–54.

Monasterolo, I. (2020), “Climate Change and the Financial System, An-nual Review of Environment and Resources”, Annual Review of Resource Economics 12(1), 299–320.

Kittner, N., F. Lill, and D.M. Kammen (2017), “Energy Storage De-ployment and Innovation for the Clean Energy Transition,” Nature Energy 2(9).

TEG (EU Technical Expert Group on Sustainable Finance) (2019a), Financing a Sustainable European Economy, Using the Taxonomy – Sup-plementary Report 2019.

TEG (EU Technical Expert Group on Sustainable Finance) (2019b), Financing a Sustainable European Economy, Taxonomy – Technical Report.

CONTENT

34 CESifo Forum 3 / 2021 May Volume 22

RESEARCH REPORT

Since the onset of the pandemic, parameterized ep-idemiological SIR (susceptible, infected and recov-ered) models have been a popular tool for analyzing the disease dynamics (Anderson et al. 2020; Atkeson 2020; Stock 2020). These models can be used to shed light on the effectiveness of physical distancing and other public health measures in containing a second wave of infections (Ferguson et al. 2020; Matrajt and Leung 2020; Davies et al. 2020; Hornstein 2020). SIR models rely on several parameters (to quantify the impact of physical distancing on the basic reproduc-tion or ‘R’ number, for instance), so their insights are only as good as these parameters are accurate. So far, however, most modelers have relied on past (mostly flu) epidemics to calibrate important pa-rameters, while the current pandemic likely differs in important ways.

This study contributes to a burgeoning litera-ture that seeks to quantify the impact of government interventions on disease progression and mobility by employing reduced-form econometric estimates of the Covid-19 pandemic itself. This literature has already shown that stricter lockdown policies go in tandem with a reduction in Covid-19-related deaths (Conyon et al. 2020). It has found strong evi-dence that banning mass gatherings is one of the most effective ways of taming the spread of the virus (Ahammer et al. 2020; Hunter et al. 2020; Weber 2020). Similarly, air travel restrictions are found to be effective, especially those imposed on interna-tional flights and at the early stages of the pandemic (Hubert 2020; Keita 2020, Leffler et al. 2020). Stay-at-home requirements and workplace closures can also curb the propagation of the disease (Deb et al. 2020; Hunter et al. 2020; Weber 2020), as can the use of face masks (Hatzius et al. 2020; Leffler et al. 2020; Mitze et al. 2020). Nevertheless, recent empirical literature has said little about the importance of testing and contact tracing policies (despite their prominence in SIR models) and the protection of the elderly population.

The rest of the paper is structured as follows. Section 2 presents the drivers of the reproduction rate, while Section 3 describes the determinants of mobility. Section 4 offers a scenario analysis and Sec-tion 5 provides some concluding remarks.

THE IMPACT OF POLICIES ON THE REPRODUC-TION RATE

Confinement Policies

In estimation, the coefficients on five con-tainment policies workplace closures, re-

strictions on gatherings, stay-at-home re-quirements, international travel controls and school closures1 are found to have a

statistically significant effect in reducing R

1 Containment measures are drawn from the Oxford Blavatnik School of Government (Hale et al. 2020) and are variously scored 0 to 2, 0 to 3, or 0 to 4. In this paper, four different levels of stringency are included as distinct dum-my variables (taking the value of zero or one) as there is no reason to expect a policy with a stringency value of 3 to have triple the effect of a policy with a value of 1). Data on mask wearing and protection of the elderly is constructed by the authors using textual extraction from various data-bases. The working paper version of this paper provides details on the data used for the estimations and some of the modelling choices: Égert et al. (2020).

Balázs Égert, Yvan Guillemette, Fabrice Murtin and David Turner

Walking the Tightrope: Avoiding a Lockdown While Containing the Virus*

is Senior Economist at the OECD and affiliated with EconomiX at the University of Paris X Nanterre and CESifo. His main research interests are international eco-nomics and finance, macro and monetary economics as well as development economics.

is Deputy Head of the Macro-economic Analysis Division at the OECD Economics Department. His main work area is estimating potential output for the OECD Economic Outlook, as well as constructing long-run macroeco-nomic and fiscal scenarios.

Balázs Égert Yvan Guillemette

This paper finds that although containment policies can suc-cessfully reduce the spread of the virus, they can also have a substantial impact on reducing mobility and economic activ-ity. It also shows that testing, combined with effective contact tracing, is crucial in reducing the spread of the virus, espe-cially at relatively low levels of infection, and that mask-wear-ing and the protection of the elderly population in general and those in care homes in particular, might play an important role in combatting the virus while minimizing economic costs.

ABSTRACT

CONTENT

35CESifo Forum 3 / 2021 May Volume 22

RESEARCH REPORT

(Table 1).2,3 The coefficient on school closures has the largest effect of any containment policies, but there is a degree of collinearity between school closures, stay-at-home re-quirements and workplace closures, which arises because such containment policies have often been imposed at the same time.

Further testing suggests that while the sum of the coefficients on these three con-tainment variables is a robust indication of the effect of a combined package, the coefficient on any one of them is less re-liable, as it is sensitive to the exclusion of other variables (for this reason, the com-bined effect of the three policies is given in Figure 5 under the heading of ‘Typical lockdown’, rather than showing each in-dividually). Similarly, the absence in the equa-tion of any role for the closure of public events is likely related to its overlap with restrictions on the size of gatherings, which is included. The com-bined effect of applying all containment polices suggests that from an initial R0 value of about 3, a complete package of containment measures would nearly halve the reproduction number.

An interesting finding is that the impact differs substantially across countries, in that workplace clo-sures have a considerably larger negative effect on R in high-income countries as compared to other coun-tries. One possible reason is that workplace closures can be enforced more effectively in high-income coun-tries, where workers, more likely to be covered by social insurance, may be less tempted to circumvent them. This finding is mirrored in the mobility equa-tion, according to which workplace closures have a larger impact in advanced economies. Conversely, while stay-at-home requirements are found to re-duce R to a greater extent in advanced economies than in less developed countries, this is not the case for mobility. The lower effectiveness of stay-at-home requirements in less advanced economies may be attributable to larger household sizes and smaller living spaces.

An important feature of these results is that the full R reduction is often achieved well before the max-imum level of stringency is reached; for example, a stringency score of 2 on the workplace closure var-iable reduces R, but no additional effect on R can be detected from a further increase in the degree of stringency. The combined effect of applying all containment polices suggests that from an initial R0 value of about 3, a complete package of containment

2 The reproduction rate is estimated using the code developed by Systrom (2020), who extends the static Bayesian approach of Betten-court and Ribeiro (2008).3 An important feature of the estimated equation explaining R is that the preferred functional form for the dependent variable is loga-rithmic; a formal test decisively rejects a linear form in favour of a logarithmic one. This implies that any policy intervention will have a larger effect when R is initially high than when it is low, and under-lines the merit of early policy interventions.

measures would nearly halve the reproduction num-ber (Figure 1).

Test and Trace Policies

Results suggest that test and trace policies can re-duce the spread of the virus (Table 1). The most com -prehensive form of test and trace policies4 are more than 2½ times as effective in reducing R than more limited forms. Test and trace polices are most effec-tive when the infection rate is not too high (which in estimation is taken to be less than 10 new daily cases per million population, a rate which was well exceeded by many countries in March and April 2020), a rather unsurprising finding given the difficulties of tracking down all contact persons in a timely manner when the system is overwhelmed with new cases.

4 Most comprehensive testing is defined as open public testing (e.g., ‘drive through’ testing available to asymptomatic people). Most comprehensive tracing is defined as comprehensive contact tracing done for all identified cases.

-0.3076 -0.2798

-0.0883-0.0540

-0.0629 -0.1840

-0.1022

-0.1370

-0.8

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

0

Containment policies Public health policies

Most comprehensive test, trace and isolate policies

Testing in care homes

Restricting visits to care homes

Recommend elderly stay-at-home

Restricting gatherings

International travel controls

Typical lockdown1

Note: This chart decomposes the effect on (logged) R of the various containment policies (bars on left) and public health policies (bars on right) according to the regressions in Table 5.

The effects of school closures (1 ≥2), stay-at-home requirements (≥1) and workplace closures (≥2) have been combined into a single segment labelled 'Typical lockdown'. This is because such policies have often been imposed at the same time and, as discussed in the main text, because multi-collinearity means that the sum of the coefficients on these three containment variables is more reliable than any of the individual coefficients.Source: Authors’ calculations.

Effect of Containment Policies and Public Health Policies on (Logged) R

© ifo Institute

(Logged) R

Mandatory mask-wearing

Figure 1

is Head of Research at the OECD Center on Well-Being, Inclusion, Sustainability, and Equal Oppor-tunities (WISE). His research fo-cuses on long-term economic de-velopment, labor market dynam-ics and well-being economics.

is Head of the Macroeconomic Analysis Division at the OECD. His main responsibilities currently include overseeing the ADB macroeconomic database as well as overseeing analysis of poten-tial output, modelling the effects of structural reforms and long-term projections for the global economy.

Fabrice Murtin David Turner

CONTENT

36 CESifo Forum 3 / 2021 May Volume 22

RESEARCH REPORT

Overall, the impact of the most effective test and trace regime on reducing R in an environment of low daily infection is estimated to be greater than any other public health intervention and 2–3 times more effective than most individual containment measures (Figure 1). Variant regressions show that isolating the contacts of people who are sick or who tested pos-itive with the virus has a non-trivial effect on R and substantially enhances the effectiveness of test and trace policies.

Protecting the Elderly

Empirical analysis provides strong evidence that policies can play an important role in shielding the elderly population. Stepping up the testing of resi-dents and staff in long-term care facilities is found

to correlate negatively with the transmission of the virus. Restricting visitor access to these establish-ments goes in tandem with lower reproduction rates. Furthermore, general stay-at-home recommenda-tions for the elderly appears to be associated with fewer infections (Table 1). The combined effect of these polices on reducing R is estimated to exceed the effect of most individual containment measures (Figure 1).

Mask Wearing

Results show a sizeable and fairly robust negative effect on R from the introduction of mandatory mask wearing in all closed public spaces (Table 1), although other results (not reported) suggest that extending mask wearing obligations to the outdoors does not appear to add much to reducing the reproduction rate.

Awareness of the Virus and Towards Herd Immunity

Both the national and global daily death rates are included to proxy for general awareness of the vi-rus prompting more cautious behavior, for example voluntar y physical distancing and increased hand-washing. The importance of these variables is that they proxy for changes in behavior that are likely to be engendered regardless of government mandated restrictions. Total national deaths attributed to the virus expressed as a share of the population are also included separately as a proxy for the share of the population that has been infected, with the expec-tation of a negative coefficient; as the share of the population that has been infected rises (and presum-ably becomes immune), the speed with which the vi-rus spreads will be reduced. Death rate variables are statistically significant with the expected negative sign, and their magnitudes imply that they can play an important role in the evolution of R.

THE IMPACT OF CONTAINMENT POLICIES ON MOBILITY

Putting in place containment and isolation policies hinders the free daily movement of people. Empirical results suggest that seven of the eight categories of containment policy published by Google have a neg-ative effect on mobility, based on the movement of people with Android-based smartphones (Table 2).5

Unlike in the equation for R, there is a clearer ranking in coefficients, so that a more stringent application of a particular policy tends to more greatly reduce mo-bility. For example, the most severe form of workplace closure (with a score of 3) has a nine times greater ef-

5 The failure to detect any effect from restrictions on gatherings is likely related to its close correlation with the policy to cancel public events.

Table 1

Drivers of the Reproduction RateSample period: 1 January to 17 August 2020

Dependent variable: ln(R)

Constant 1.0947**

Containment policies

Stay-at-home requirement (≥1) – 0.0536**

Workplace closures (=1) – 0.0614**

Workplace closures (≥2) – 0.0767**

School closures (≥2) – 0.1773**

Restrictions on gatherings (=2) – 0.0393**

Restrictions on gatherings (≥3) – 0.0883**

International travel controls (≥1) – 0.0629**

Test and Trace policies

Test=1 or 2, Trace =1 or 2 – 0.1110**

Test=3, Trace=1 – 0.1364**

Test=3, Trace=2 – 0.2185**

All Test & Trace combinations when deaths < 10 per million – 0.0613**

Policies protecting the elderly

Testing in care homes (=2) – 0.0540**

Restricting visits to care homes (≥1) – 0.1840**

Recommending elderly to stay at home – 0.1022**

Other non­containment policies

Mandatory mask wearing indoors – 0.1370**

Death rates (per million population)

Daily national – 0.0358**

Daily global – 0.3637**

Total national – 0.0007**

Adjusted R-squared 0.597

Daily observations 17,624

Countries covered 147

Country fixed effects Yes

Note: For details of the construction of data on R see Annex A. The policy variables are based on the variables described in Tables 1 to 3 in the main text, but re-normalized to (0, 1) dummy variables as described in the main text. The notation in brackets (=n) given after a containment policy variable denotes that the dummy variable is assigned a 1 if the original score for that policy was equal to n, whereas the notation (≥n) denotes that the dummy variable is assigned a 1 if the original score for that policy was greater than or equal to n. ** denotes statistical significance at the 5% level, based on heteroscedasticity-robust standard errors.

Source: Authors’ calculations.

CONTENT

37CESifo Forum 3 / 2021 May Volume 22

RESEARCH REPORT

fect on mobility of the mildest form (with a score of 1). These findings suggest that moving to more stringent forms of workplace closure, stay-at-home require-ments and school closure has larger negative effects on mobility and hence economic activity, although it is difficult to detect any corresponding benefit from further reductions in R.

For policies such as the cancellation of public events and travel restrictions, the most limited ap-plication of the policy has no significant effect on mobility. Applying all containment policies in their most severe forms would reduce mobility by more than half, relative to normal, with 50 percent of this reduction accounted for by workplace closures and stay-at-home requirements.

Alternative estimations explore the effect of mask-wearing on mobility. Positive coefficient esti-mates suggest that mandating mask wearing in pub-lic transport and shops increases mobility. Similarly, more extensive testing and the isolation of contact persons are found to encourage mobility (Table B.6 in Annex B), possibly by reducing concerns about infection.

The national daily death rate from the virus is again included to proxy general awareness of the virus and its effect in voluntarily reducing mobility due to an increase in natural caution. A national daily death rate running at around 15 per million – similar to the rate experienced by some major OECD countries go-ing into the lockdown in March 2020 – would reduce mobility by 10 percent, independently of any govern-ment-mandated polices.

SCENARIO ANALYSIS

A number of stylized scenarios are constructed using the estimated equations for R and mobility (Figure 2, Table 3). For a typical country, at the first outbreak of the virus, the initial reproduction number R0 is esti-mated to be about 3, and mobility is normal before the impact of the virus on the economy is felt (repre-sented by the red triangle at the top right-hand cor-ner of Figure 2). Even before the implementation of government-mandated measures, awareness of the seriousness of the virus (represented by the daily death rate) is likely to reduce mobility and foster more cautious behavior, leading to a fall in R, although it remains well above 1.0 (indicated by the red triangle at ‘Pre­lockdown + natural caution’ in Figure 2, which is calibrated on the daily death rates of a number of major OECD economies just prior to lockdown).

Once the number of daily infections is high (here proxied by the high national daily death rate), the im-plementation of a wide range of containment meas-ures is essential to contain the spread of the virus. In the scenarios considered here, the implementation of full lockdown (FLD) measures, accompanied by a limited test-and-trace regime, reduces R to close to 1.0, but at the cost of a sharp fall in mobility (rep-

resented by the blue squares in Figure 2). The deb-gree of stringency with which lockdown measures

Table 2

Drivers of MobilitySample period: 1 January to 17 August 2020

Dependent variable: Mobility

Constant 1.0241**

Containment policies

Stay-at-home requirement (=1) – 0.0240**

Stay-at-home requirement (=2) – 0.0668**

Stay-at-home requirement (=3) – 0.1252**

Workplace closures (=1) – 0.0216**

Workplace closures (=2) – 0.0491**

Workplace closures (=3) – 0.1980**

School closures (=2) – 0.0237**

School closures (=3) – 0.1098**

Canceling public events (=2) – 0.0369**

Restrictions on internal movement (=2) – 0.0220**

International travel controls (=4) – 0.0554**

Close public transport (=1) – 0.0439**

Close public transport (=2) – 0.0650**

Death rate (per million population)

Daily national – 0.0066**

Adjusted R-squared 0.759

Daily observations 22,741

Countries covered 128

Country fixed effects Yes

Note: Mobility data are made available by Google, based on the movement of people with ‘location history’ turned on in their smartphone settings. The index used here measures the change in mobility from a same-day-of-the-week average in January and early February, such that normality would suggest an index of 1.0. The containment policy variables are based on those given in Table 1, but re-normalized to (0, 1) dummy variables as described in the main text. The notation in brackets (=n) given after a containment policy variable denotes that the dummy variable is assigned a 1 if the original score for that policy was equal to n, whereas the notation (≥n) denotes that the dummy variable is assigned a 1 if the original score for that policy was greater than or equal to n. ** denotes statistical significance at the 5% level.

Source: Authors’ calculations.

0

20

40

60

80

100

120

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5

R0 pre-lockdown

Pre-lockdown & natural caution

Partial LD + full health measures

No LD + full health measures

No LD + limited health measures

Severe full lockdown

Stylized Scenarios: From the First Outbreak of the Virus, Through Lockdown (LD) and Exit

Note: The points represent scenarios, each of which are generated from consistent combinations of the equations for R and mobility, using assumptions for the explanatory variables that are presented in Table 3. Red triangles denote the situation at the start of the virus outbreak, blue squares indicate the situation following full lock-down policies, and green circles represent various exit scenarios.Source: Authors’ calculations. © ifo Institute

Reproduction number, R

Mobility (% relative to normal)

Full lockdown

Partial LD + limited health measures

Figure 2

CONTENT

38 CESifo Forum 3 / 2021 May Volume 22

RESEARCH REPORT

Tabl

e 3

Scen

ario

Ass

umpt

ions

and

Out

com

es fo

r R a

nd M

obili

ty

Scen

ario

Cont

ainm

ent

mea

sure

sTe

stin

g, tr

acin

g &

isol

atio

nO

ther

pub

lic h

ealth

pol

icie

sDa

ily n

atio

nal

deat

hsDa

ily g

loba

l de

aths

Tota

l nat

iona

l de

aths

RM

obili

ty in

dex

(1.0

0 =

norm

al)

Exte

nsiv

eLi

mite

d M

ask-

wea

ring

Test

ing

in c

are

hom

esEl

derl

y st

ay a

t ho

me

Bann

ing

of

care

hom

e vi

sits

per m

illio

n of

pop

ulat

ion

Pre-

lo

ckdo

wn

R0N

one

--

--

--

00

03

1

Nat

ural

cau

tion

Non

e-

--

--

-5.

00.

850

1.81

0.96

7

Lock

dow

n

Full

lock

dow

n (F

LD)

Com

preh

ensi

ve-

√-

--

-5.

00.

850

1.02

0.56

2

Seve

re F

LD

Com

preh

ensi

ve &

se

vere

-√

--

--

5.0

0.8

501.

020.

3547

Exit

from

lock

dow

n

Part

ial L

D &

full

heal

th m

easu

res

Bann

ing

of la

rge

publ

ic e

vent

s,

rest

rict g

athe

rings

, qu

aran

tine

inte

rnat

iona

l tr

avel

lers

√-

√√

√√

1.0

0.7

300

0.73

0.83

52

Part

ial L

D &

limite

d he

alth

mea

sure

s Ba

nnin

g of

larg

e pu

blic

eve

nts,

re

stric

t gat

herin

gs,

quar

antin

e in

tern

atio

nal

trav

elle

rs

-√

√-

--

1.0

0.7

300

1.22

0.83

52

No

LD &

full

heal

th

mea

sure

sN

one

√-

√√

√√

1.0

0.7

300

0.85

0.99

34

No

LD &

lim

ited

heal

th m

easu

res

Non

e-

√√

--

-1.

00.

730

01.

420.

9934

Not

e: T

he a

ssum

ptio

ns h

ere

corr

espo

nd to

the

scen

ario

s illu

stra

ted

in F

igur

e 2.

Sour

ce: A

utho

rs’ c

alcu

latio

ns.

CONTENT

39CESifo Forum 3 / 2021 May Volume 22

RESEARCH REPORT

are applied will determine the extent of the fall in mobility, with two scenarios considered here. The first assumes that containment policies are applied with a degree of stringency which is typical of that followed by countries in March/April 2020 (corre-sponding to the median country), while the second assumes that all containment policies are applied to their maximum possible degree of stringency. Mobil-ity falls by more than 40 percent in the former case and by more than 60 percent in the latter; however, the estimation results suggest there is little addi-tional benefit from maximizing the degree of strin-gency of containment policies in terms of lowering R (particularly with regard to workplace closures or stay-at-home requirements).

Even in the absence of further policy changes, the reproduction number will evolve during lockdown as the numbers of infections and deaths change. The fall in the daily death rate may tend to lower natu-ral caution and so lead to some increase in R and mobility; on the other hand, as the total number of individuals who have already been infected and have acquired immunity rises, then this will tend to lower R. The estimation results and particular calibrations used in constructing these scenarios suggest these two effects roughly cancel each other out.

A number of strategies for avoiding a full lock-down are considered (represented by the green circles in Figure 2). The basic issue facing policymakers is how to prevent the need for a full set of containment policies while bringing or keeping R under control. The estimation results (Table 1) suggest that the ims-plementation of a comprehensive test and trace pol-icy along with a package of additional public health measures would more than compensate for the re-moval of lockdown policies, such that their successful implementation would see a return to near normality of mobility, with R remaining below 1 (as represented by the green circle labelled ‘No LD + full health meas­ures’ in Figure 2).

An even more decisive reduction in R below 1 might be achieved if comprehensive public health measures were accompanied by some containment policies being maintained (here assuming that restric-tions on large public events, large public gatherings and international travel remain), although it would come at some cost to mobility (‘Partial LD + full health measures’ in Figure 2).

In practice, as the experience of several countries is showing, implementing a full range of public health policies and a comprehensive test and trace regime may be difficult, especially once the daily infection rate has begun to rise. Variant scenarios with ‘limited health measures’ assume only a limited test-and-trace regime along with mandated mask-wearing in indoor public places, but no other public health policies tar-geted at the elderly or care homes. Such a combina-tion of policies accompanied by a full relaxation of lockdown measures might see mobility initially return

to just below normal levels (assuming the daily death rate has previously been reduced by lockdown), but R will likely increase well above 1.0 (represented by the scenario labelled ‘No LD + limited health measures’ in Figure 6). However, this situation would not be a stable equilibrium, as with R above 1.0 there would be a subsequent pick-up in infections and deaths, which in turn would further reduce mobility, regardless of further government action.

A limited set of health measures accompanied by the same limited containment policies being main-tained would come at a more immediate cost to mo-bility but bring R down further, although in the sce-nario considered here it would still remain above 1 (‘Partial LD + limited health measures’), and so would not represent a sustainable situation.

CONCLUSION

This study investigated the impact of different gov-ernment interventions on both the reproduction rate of the virus, R, and mobility, as a proxy for economic activity. The empirical results inform a number of sce-narios, in which the epidemic/economic trade-off of different policy packages is assessed.

First, when the daily infection rate is high, a com-prehensive combination of containment policies is needed to reduce the spread of the virus, although these are likely to severely reduce mobility and eco-nomic activity.

Second, once the daily infection rate has been lowered, test-and-trace policies represent a better alternative for controlling the virus, because they have no significant adverse impact on mobility or economic activity. Testing is found to be more effec-tive in reducing R if accompanied by comprehensive contact tracing and is most effective in a low-infec-tion environment, because contact tracing becomes increasingly difficult with higher levels of new daily infections (OECD 2020). Specific testing in care homes is also important to control the spread of the virus.

Third, other public health policies can also con-tribute to restraining the spread of the virus, including mandating mask-wearing in public indoor environ-ments, restricting visits to care homes, and stay-at-home recommendations for the elderly population.6

Even with a comprehensive test-and-trace regime and supporting public health policies, there may be a need to resort to selective containment measures. These should prioritize restrictions on large gather-ings and international travel. Where there are local-ized outbreaks of the virus, then targeted lockdown measures are appropriate.

Finally, vaccination is likely to become the most important policy with which to combat the spread of 6 These findings are in line with and complement Acemoglu et al. (2020), who show in a multi-group SIR framework that the trade-off between mortality rates from the virus and economic damage can be attenuated if interventions are targeted on the most vulnerable indi-viduals.

CONTENT

40 CESifo Forum 3 / 2021 May Volume 22

RESEARCH REPORT

the virus in the coming year(s). There already seems to be clear evidence that the vaccination program in Israel is reducing the spread of the virus there (Ross-man et al. 2021), but as yet it is difficult to identify evidence of vaccinations in multi-country estimations due to the lack of available data. However, any future work should have a special focus on the impact of vaccination programs.

REFERENCES Acemoglu, D., V. Chernozhukov, I. Werning, and M. D. Whinston (2020), “Optimal Targeted Lockdowns in a Multigroup SIR Model”, NBER Working Paper No. 27102.

Ahammer, A., M. Halla, and M. Lackner (2020), “Mass Gathering Contrib-uted to Early COVID-19 Spread: Evidence from US Sports”, CEPR Covid Economics 30.

Anderson, R.M., H. Heesterbeek, D. Klinkenberg and T.D. Hollingsworth (2020), “How Will Country-based Mitigation Measures Influence the Course of the COVID-19 Epidemic?”, The Lancet 395(10228), 931–934.

Atkeson, A. (2020), “What Will be the Economic Impact of COVID-19 in the US? Estimate of Disease Scenario”, NBER Working Paper No. 26867.

Bettencourt L.M.A. and R.M. Ribeiro (2008), “Real Time Bayesian Estima-tion of the Epidemic Potential of Emerging Infectious Diseases”, PLOS ONE 3(5).

Conyon, M.J., L. He, and S. Thomsen (2020), “Lockdowns and Covid-19 Deaths in Scandinavia”, CEPR Covid Economics 26.

Davies, N.G., A.J. Kucharski, R.M. Eggo, A. Gimma, and W.J. Edmunds (2020), “Effects of Non-pharmaceutical Interventions on Covid-19 Cases, Death and Demand for Hospital Services in the UK: A Modelling Study”, The Lancet Public Health 5(7).

Deb, P., D. Furceri, J. D. Ostry, and N. Tawk (2020), “The Effect of Containment Measures on the Covid-19 Pandemic”, CEPR Covid Economics 19.

Égert, B., Y. Guillemette, F. Murtin, and D. Turner (2020), “Walking the Tightrope: Avoiding a Lockdown While Containing the Virus”, OECD Economics Department Working Paper No. 1633.

Ferguson et al. (2020), “Impact of Non-Pharmaceutical Interventions (NPIs) to Reduce COVID-19 Mortality and Healthcare Demand”, Imperial College COVID-19 Response Team, London.

Hale, T. et al. (2020), “Oxford COVID-19 Government Response Tracker”, Blavatnik School of Government, Oxford University.

Hatzius, J., D. Struyven, and I. Rosenberg (2020), “Face Masks and GDP”, Goldman Sachs Global Economics Analyst, 29 June.

Hornstein, A. (2020), “Social Distancing, Quarantine, Contact Tracing and Testing: Implications of an Augmented SEIR Model”, CEPR Covid Economics 18.

Hubert, O. (2020), “Spacial Distancing: Air Traffic, Covid-19 Propagation and the Cost of Efficiency of Air Travel Restrictions”, CEPR Covid Eco­nomics 24.

Hunter, P.R., F.J. Colón-González, J. Brainard, and S. Rushton (2020), “Impact of Non-pharmaceutical Interventions Against COVID-19 in Eu-rope: A Quasi-experimental Study”, mimeo.

Keita, S. (2020), “Air Passenger Mobility, Travel Restrictions and the Transmission of the Covid-19 Pandemic between Countries”, CEPR Covid Economics 9.

Leffler, G.T., E. Ing, J.D. Lykins, M.C. Hogan, C.A. McKeown, and A. Grzy-bowski (2020), “Association of Country-wide Coronavirus Mortality with Demographics, Testing, Lockdowns, and Public Wearing of Masks”, mimeo.

Matrajt, L. and T. Leung (2020), “Evaluating the Effectiveness of Social Distancing Interventions to Delay or Flatten the Epidemic Curve of Coro-navirus Disease”, Emerging Infectious Diseases 26(8).

Mitze, T., R. Kosfeld, J. Rode, and K. Wälde (2020), “Face Masks Consid-erably Reduce Covid-19 Cases in Germany: A Synthetic Control Method Approach”, CEPR Covid Economics 27.

OECD (2020), Testing for COVID­19: A Way to Lift Confinement Re­strictions, https://www.oecd.org/coronavirus/policy-responses/testing-for-covid-19-a-way-to-lift-confinement-restrictions-89756248/.

Rossman, H. et al. (2021), “Patterns of COVID-19 Pandemic Dynamics Following Deployment of a Broad National Immunization Program”, medRxiv.

Stock, J.H. (2020), “Data Gaps and the Policy Response to the Novel Coronavirus”, NBER Working Paper No. 26902.

Systrom, K. (2020), Estimating COVID­19’s Rt in Real­Time, IPython Notebook.

Weber, E. (2020), “Which Measures Flattened the Curve in Germany?”, CEPR Covid Economics 24.

CONTENT

41CESifo Forum 3 / 2021 May Volume 22

REFORM MODEL

To implement the Green Deal, the EU is currently pre-paring a major overhaul of its climate policy frame-work. The Green Deal is the bloc’s masterplan for tran-sitioning to a sustainable economy in this decade and beyond. As a first step, the EU Council agreed in De-cember 2020 to increase its long-term climate targets, and now aims for a 55 percent reduction in green-house gas (GHG) emissions relative to 1990 levels by 2030, and GHG neutrality by 2050. The second step is to open up the entire climate and energy package for revision and subsequent reform. In summer 2021, the European Commission (EC) will publish its proposals for a revision of the full range of climate measures, most crucially the Emissions Trading System Direc-tive (EU-ETS Directive) and the Effort Sharing Reg-ulation (ESR). Although the EU’s ambitions require urgent action, decisions must be well informed and economically prudent – especially since this set of reforms will set the direction of policy architecture for decades to come.

INTRODUCTION

The impact assessment1 (IA) accompanying the EC’s Climate Target Plan 2030 already alludes to policy-makers’ preference for a broad policy mix. The IA ana-lyzes various future policy architecture scenarios, of which the three main ones describe differing roles for carbon pricing and other policies and measures (see Table 1). In the “CPRICE” scenario, a new econ-omy-wide ETS would be the centerpiece of the new policy architecture, eschewing regulation in favor of a market-based approach. In contrast, the “REG” sce-nario is based on a high intensification of policies and measures, implying a regulatory philosophy of “com-mand and control”. The “MIX” scenario, as the name suggests, sits in between, with no specific regulatory philosophy, and is expected to meet the least polit-ical resistance to implementation. This makes it the most likely path for future reforms (Knodt et al. 2020).

1 https://ec.europa.eu/clima/sites/clima/files/eu-climate-action/docs/impact_en.pdf

A major challenge in designing policy mixes, however, is to ensure coherence such that indi-vidual instruments complement rather than coun-teract each other. Moreover, a dedicated compliance mechanism is needed to ensure that policy targets are met, especially when the interaction between instruments is complex and their impact uncertain. If a policy mix is not appropriately designed and governed, future discretionary policy adjustments are very likely, especially if the policy mix has no clear-cut regulatory philosophy. The profound im-plications of such an approach to regulation were exposed by Kydland & Prescott (1977) in their semi-nal paper on rules vs. discretion. They showed that discretion can destabilize the market and lead to an iterative reform process that never converges. Discre-tion also implies that policymakers cannot make a

Michael Pahle and Ottmar Edenhofer*

Discretionary Intervention Destabilizes the EU Emissions Trading System: Evidence and Recommendations for a Rule-Based Cap Adjustment

The EU is currently preparing a major overhaul of its climate policy framework to deliver on the Green Deal’s new climate targets of a 55 percent cut in greenhouse gas (GHG) emis-sions relative to 1990 by 2030 and GHG neutrality by 2050. Ex-tending and strengthening the role of carbon pricing, imple-mented through the EU Emissions Trading System (EU-ETS), will play an important role in this framework. Accordingly, the design and governance of the EU-ETS will be ever more crucial. In this article, we focus both on the 2018 EU-ETS re-form as the first step on a slippery slope of increasing dis-cretionary intervention and on the upcoming reform risks reinforcing this trend. In their seminal work, Kydland and Prescott (1977) caution against such interventions, because of their ability to destabilize the market and engender recur-ring interventions. This limits the capacity of policymakers to credibly commit to long-term targets, which undermines the dynamic efficiency of intertemporal emissions trading sys-tems like the EU-ETS. To counteract this trend, we provide recommendations for rule-based adjustments to the EU-ETS.

ABSTRACT

* This article builds on research conducted under the Koperni-kus-Projekt Ariadne (FKZ 03SFK5A), funded by the German Federal Ministry of Education and Research. The funding is gratefully ac-knowledged.

CONTENT

42 CESifo Forum 3 / 2021 May Volume 22

REFORM MODEL

credible commitment to long-term climate targets.

Yet long-term commitment is fundamental to the dynamic efficiency of intertemporal cap-

and-trade programs like the EU-ETS. If credible, this commitment brings a degree of certainty re-garding future reductions in the emissions cap, thus avoiding dis-tortions in investment decisions, the hold-up problem and the in-

appropriate allocation of risk (Hepburn 2006). At the same time, regulators want the flex-ibility to be able to cope with

future circumstances, which is why they prefer discretion-ary intervention to commitment (Hasegawa and Salant 2014). However, this can result in eco-nomic instability; for instance, firms may persistently mispredict the intervention’s long-term im-pact on allowance prices, which may necessitate further correc-tive interventions in the future

(Kydland and Prescott 1977). The anticipation of future corrections might create regulatory uncer-tainty which, in turn, might further distort allowance prices (Salant and Henderson 1978; Salant 2016). Contrary to their intention, then, discretionary in-terventions in cap-and-trade programs could have the perverse effect of destabilizing the allowance market.

Against this background, we first clarify the risks and implications of such discretionary interventions for the EU-ETS. We focus on emissions trading be-cause of its relevance for EU climate policy and be-cause its dynamic efficiency can be substantially impaired by such interventions. We discuss two examples: the 2018 EU-ETS reform and the 2020 German coal phase-out, the latter being a national measure that overlaps with the EU-ETS. We then provide recommendations for the further evolution of the EU-ETS in line with a rule-based approach,

lending credibility to the EU’s long-term commit-ment by removing the specter of future discretionary intervention.

DISCRETIONARY INTERVENTION IN THE EU-ETS

Historically, the main trigger of intervention in the market has always been the level of allowance prices. From an economic perspective, it is precisely this flex-ibility of allowance prices in response to supply and demand that constitutes its efficiency. If prices are very high, this reflects a high demand or sparse sup-ply – and vice versa. Correspondingly, if the supply of allowances (i.e., the cap) is fixed, the allowance price adjusts itself to the marginal emissions abatement cost, ensuring that climate targets are met. However, the low price of allowances in the past have triggered a debate both about the proper functioning of the EU-ETS and its broader role in the policy mix (Flachsland et al. 2020). From a political perspective, this implies a risk of the program becoming insignificant, and po-litical support may wane (Borenstein 2016). On the other hand, there might also be a backlash if allow-ance prices exceed a politically acceptable level (Bo-renstein et al. 2019).

In fact, “measures in the event of excessive price fluctuations” were already included in the 2009 re-form of the EU-ETS.2 This happened as a response to the sharp rise in allowance prices in 2008. How-ever, in the subsequent years, such a rise never oc-curred again (see Figure 1), so these measures have as yet never been executed. Yet their very adoption is evidence that regulators are ready and willing to intervene in the market. This indeed happened later on, when prices dropped to levels that were far lower than expected (Ellerman, Marcantonini, and Zaklan 2016), triggering important interven-tions that adjusted supply by canceling allowances. At the same time, allowance cancelation has also been deployed at the member state level (Germany) to compensate for complementary climate policies. The following sections discuss these two interven-tions and their impacts in relation to Kydland and Prescott’s findings. 2 See Article 29a at: https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32009L0029&from=EN.

is Head of the “Climate and En-ergy Policy” working group at the Potsdam Institute for Climate Impact Research (PIK).

Michael Pahle

is Director of PIK, Director of the Mercator Research Institute on Global Commons and Climate Change (MCC), and Professor of the Economics of Climate Change at the Technical University Berlin.

Ottmar Edenhofer

Table 1

Main Policy Architecture Scenarios of the EC Impact Assessment

Increasing role of carbon pricing

REG MIX CPRICE

ETS Same sectoral scope Extension to other sectors(including buildings and transport)

Policies & measures High intensification Medium/low intensification EE* and RES**: no intensificationTransport: low intensification

ESR Same sectoral scope Same sectoral scope ESR does not apply to buildings & transport

*EE=energy efficiency; **RES=renewable energy sources.

Source: Knodt et al. (2020) based on EC.

CONTENT

43CESifo Forum 3 / 2021 May Volume 22

REFORM MODEL

Cancelation of Allowances in the 2018 EU-ETS Reform

After the first historical high in June 2008, prices in the EU-ETS declined to relatively low levels and lin-gered there for several years (see Figure 1). Diagnos-ing a systematic imbalance in supply and demand, EU regulators started to take counteractions by first introducing the “Market Stability Reserve”, a cap-neu-tral mechanism for moving allowances into and out of the market depending on the total number of al-lowances in circulation. Later, as part of the reform of the EU-ETS in preparation for the next trading phase (2021–2030)3, they adjusted the cap by means of the so-called “Linear Reduction Factor”, which progres-sively lowers the cap according to a predetermined path. At the same time, they strengthened the Market Stability Reserve by introducing the cancelation of al-lowances from 2023 onwards, which also indirectly af-fects the cap. The reform was signed into law in March 2018, and prices started to rally to levels nearing 30 EUR/t within a few months. Considering prices alone, it would appear that the reform delivered on rectifying the “diagnosed” supply–demand imbalance, as expec-tations of a lower supply in the future indeed led to an increase in prices, as economic theory suggests.

However, the central question here is not how prices responded, but how the reform affected mar-ket beliefs about the price impact of the reform and future interventions. Economic theory predicts that if market participants expect such interventions, inter-temporal price formation becomes distorted (Salant 2016). Accordingly, because Market Stability Reserve cancelation rules are very complex (Perino 2018) and their effect on prices is very difficult for the market to predict, the reform may actually have destabilized the market. Notably, the Market Stability Reserve’s complexity may have been deliberate, something which has been characterized as “smokescreen poli-tics” (Wettestad and Jevnaker 2019), that is, making the distribution of costs obscure and diffuse and the probable benefits (a higher carbon price and therefore greater auctioning revenues for member states) more specific and closer in time. This will have helped to make the reform politically feasible, but at the cost of increasing price uncertainty.

Before proceeding, it is helpful to discuss whether the reform constituted a discretionary intervention or not. The answer is ambiguous. It was discretionary in the sense that the supply-demand imbalance trig-gered the 2015 reform (as described above), and influ-enced the 2018 reform. It was rule-based in the sense that the intervention was scheduled years ahead of its planned implementation, and changes to the mar-ket design were also rule-based – unlike the one-off (discretionary) cancelation of a certain number of al-lowances, for example. Yet, as mentioned, the com-3 For more detail see: https://ec.europa.eu/clima/policies/ets/revi-sion_en

plexity of the cancelation rules makes their impact on prices very hard to predict. This is underlined by the wide-ranging estimates of the number of allowances that will be canceled from 2023, from around 1 Gt to 13 Gt (Osorio et al. 2020).

It was against this background that we tested the hypothesis in a recent analysis (Friedrich et al. 2020) that the reform prompted market participants to persistently speculate about its price impacts, thereby destabilizing the market. For this analysis we employed the following empirical strategy. First, we used time-varying regression to analyze whether the price trend can be explained by a correspond-ing development in market fundamentals. Second, we conducted a statistical test to look for a period of explosive behavior, indicating an overreaction of the market (similar to what is observed during the inflationary period of a bubble), and whether it coin-cides with the date of the reform’s adoption. Third, we modeled the episode using non-causal statistical processes to calculate the crash odds.

The results of the first two steps suggest that the reform indeed triggered a speculative overreaction about its price effects. The results of the time-vary-ing coefficient model confirm that the upward trend cannot be explained by movements in market funda-mental price drivers. However, it is picked up by the

0

10

20

30

40

50

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021

Source: Sandbag.

EU-ETS Allowance Price

Price (€/t)

© ifo Institute

Figure 1

-2

-1

0

1

2

3

4

5

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Test statistics Critical values

Note: Bootstrapped critical value series (blue) and test statistics (black). Sequences of test values above critical values indicate explosive episodes. Source: Friedrich et al. (2020).

Date Stamping of Explosive Episodes in Allowance Prices Using the BSADF (Backward Supremum Augmented Dickey-Fuller) Test

© ifo Institute

Figure 2

CONTENT

44 CESifo Forum 3 / 2021 May Volume 22

REFORM MODEL

trend component, which implies other price drivers. Correspondingly, we find evidence of explosive be-havior beginning in March 2018, coinciding exactly with the time when the reform was signed into law. Finally, the upper estimates for the one-year-ahead crash odds after October 2018 (when prices peaked) are as high as 86 percent, which further supports the interpretation that the price rally was an overreaction of the market.

While these empirical results lend themselves to different explanations, the characteristics of the reform suggest that it has indeed destabilized the market. It must be noted that a process of adapting to a new equilibrium can also appear explosive (Harvey et al. 2016), although this would not result in market destabilization. However, it is questionable whether the market is now in a stable equilibrium, given un-certainties surrounding the cancelation volume, and therefore whether this explanation is applicable. Fur-thermore, while no rock-bottom price collapse has materialized (see Figure 1), there was a (partial) col-lapse from around 25 EUR/t down to around 15 EUR/t in November 2018. The price subsequently rebounded, but this might have been due to a conflation with fur-ther policy developments – the EC published its vi-sion for a 2050 climate-neutral economy in November 2018. Moreover, in the course of 2019, it became clear that the EC would propose more stringent climate targets for 2030, implying a more stringent cap, and in December 2019, it published the EU Green Deal, which proposes a ratcheting up of long-term climate targets. This development may have counteracted further price drops.

The policy implications of these findings are as follows: While rules are preferred over discretion, it is important for the rules to be clear and transpar-ent in terms of their impact on prices. If the market cannot predict their impact, it may destabilize price formation and induce volatility. With the EU-ETS in-creasingly becoming a financial market, such volatility might be compounded if noise traders are attracted (De Long et al. 1990). In fact, media reports suggest that the 2018 reform encouraged many financial players, such as investment banks and hedge funds, to enter the market. The findings also question the stability of the rule itself. Even if policy makers limit themselves to only adjusting the long-run cap, this could still induce substantial price volatility in the short term. If prices become politically unacceptable (too low or too high) as a result, policymakers may be unwilling to uphold the rule and intervene (see above). We return to this issue in the final section.

Alleviating the Waterbed Effect of the 2030 German Coal Phase-Out

The persistently low EU-ETS allowance prices through-out 2017 also had important implications for interven-tions at the member state level. Notably, Germany

was bound to fall short of its ambitious national cli-mate targets, not least because it failed to reduce its coal-based emissions. As a result, the so-called “Coal Commission”4 was established to propose measures that would enable the country to meet its 2030 target for the electricity sector. The commission ultimately proposed a timeline for phasing out coal capacity, similar to the nuclear phase-out – rejecting the adop-tion of a national or EU-wide carbon price floor (Ed-enhofer and Pahle 2019).

The specific policy aside, any national measure to phase out coal will necessarily overlap with the EU-ETS, leading to the so-called waterbed effect: with a given cap, policy-related emission reductions are neutralized by increased emissions elsewhere in the system. In anticipation of this effect, the 2018 EU-ETS reform included a provision (Article 12(4)) for the uni-lateral de-facto cancelation of allowances correspond-ing to the additional emissions reductions associated with Germany’s coal phase-out. The cancelation was adopted as part of the Coal Phase-Out Act, which passed into German law in 2020.

As with the case of the Market Stability Reserve, this national cancelation mechanism can also be seen as a discretionary intervention. But an important dif-ference is that the “rule” for determining the number of allowances to be canceled is even more complex, with the volume of allowances to be canceled equal-ing the additional emission reductions from the coal phase-out (compared to a counterfactual market de-velopment without coal phase-out), net of Market Sta­bility Reserve cancelation. Both emissions estimates used in this calculation are highly uncertain, which in turn similarly renders the volume of allowances to be canceled very uncertain. What is more, overlap-ping policies may reduce the Market Stability Reserve cancelation, leading to a “green paradox” in the EU-ETS (Gerlagh, Heijmans, and Rosendahl forthcoming), whereby anticipating a drop in future allowance de-mand due to complementary policies could depress prices in the short term, reducing the total number of allowances in circulation and, in turn, the number of cancelations through the Market Stability Reserve. This could have the perverse effect that national pol-icies may even increase net emissions. Indeed, the model-based analysis by Pahle et al. (2019) confirm that the German coal phase-out has the potential to trigger this effect (see Figure 3), reducing Market Sta­bility Reserve cancelations by 164 Mt.

Here again, the (national) cancelation mechanism implies considerable price uncertainty. Notably, deter-mining the national cancelation volume depends on Market Stability Reserve cancelations and vice versa. Accordingly, uncertainties on both sides are amplified

4 Officially named the “Commission on Growth, Structural Change and Employment”, its members were tasked, inter alia, with recom-mending measures to achieve the 2030 target in the energy sector through reducing coal-based power generation. See: https://www.bmu.de/fileadmin/Daten_BMU/Download_PDF/Klimaschutz/ein-setzungsbeschluss_kohlekommission_en_bf.pdf.

CONTENT

45CESifo Forum 3 / 2021 May Volume 22

REFORM MODEL

by this interaction. This would be exacerbated if na-tional cancelations became more widespread in the future to compensate for other overlapping national measures, whether in Germany or other member states (Pahle 2020). This also underlines the coordi-nation challenges that arise from interventions at dif-ferent levels of government. In summary, by resorting to complex cancelation rules – nationally and through the Market Stability Reserve – the 2018 reform has set the EU-ETS on a path of more and more (discretion-ary) intervention.

A WAY FORWARD

The upcoming reforms of the EU-ETS in the context of the Green Deal are bound to reinforce the path taken with the 2018 reform, but reversing this trend is essential and should be prioritized. The crucial question is how best to balance flexibility (to adapt to new circumstances and information) with market stability (and thus the dynamic efficiency of the EU-ETS). Lessons from the US Clean Air Act emphasize that such adaptability is indeed a key aspect of suc-cessful climate polices (Carlson and Burtraw 2019).

The first-best approach, from an economic per-spective, would be to adjust the cap at regular inter-vals in a way that best balances benefits and costs (Hepburn 2006). Calling this a “structured discretion” approach, Aldy (2020) recently proposed such an up-dating process for a potential new US carbon tax, in which the US president would submit a carbon tax adjustment recommendation to Congress every five years, based on government agency reviews of the environmental, economic, and multilateral conditions related to climate change. It would thus incorporate flexibility for adapting to new learning and changing market conditions. The general considerations could, in principle, also be applied to adjusting the EU-ETS cap. However, a major difference between adjusting taxes and caps is that prices fluctuate after the cap is

adjusted. If the time span between cap adjustments is large, and uncertainty about future adjustment is high, considerable price uncertainty (volatility) may result. As described above, this can trigger govern-ment intervention if prices exceed or undercut polit-ically acceptable levels.

The risk of such interventions can be mitigated by a clear and transparent combined price floor and ceil-ing (or price collar). In this way, regulators can credi-bly commit to the levels at which they will intervene, assuming that such intervention is unavoidable (i.e., taking a second-best approach). A price collar would thus address the stability problem and the associated political risks. A price floor at the EU level could also help solve the coordination problem by preventing policies at the member state level from inducing the waterbed effect. In other words, a price collar ensures that prices are driven by market fundamentals alone, rather than being distorted by expectations about future regulatory interventions. For this reason, it is a central pillar of any proposals for introducing and strengthening carbon pricing in Germany (Edenhofer et al. 2019) and at the EU level (Edenhofer et al. 2021). For the EU-ETS, this would require the transformation of the Market Stability Reserve into a price-based sta-bility mechanism (Perino et al. 2021). If, as Edenhofer (2021) proposes, a parallel ETS covering the building and transport sectors were set up as an intermedi-ate step toward an integrated EU-ETS, a price collar should be implemented in this new system from the outset.

The final question concerns who should deter-mine the updates and on what basis. Kydland and Prescott’s work had a profound impact in raising pol-icymakers’ awareness of the crucial role played by credible long-term commitments in monetary pol-icy, which led to the creation of independent central banks. In the same vein, a carbon bank could be es-tablished on a similar basis to strengthen the EU’s climate policy commitment. Edenhofer et al. (2021)

-80

-60

-40

-20

0

20

40

60

2020 2025 2030 2035 2040 2045 2050

Mt CO2/a Mt CO2/aChange of emissions through coal phase out

Electricity Germany MSR cancellation

Electricity other ETS* Industry ETS**

Net change

164

-1500

-1000

-500

0

500

1000

1500Cumulative over time

*All countries regulated under ETS w/o Germany **All countries regulated under ETSNote: Red bars indicated reduction in German electricity sector. Blue and dark blue bars indicate emissions increases in other countries/sectors due to the waterbed effect. Dark red rectangles indicate changes in MSR (Market Stability Reserve) cancelations.Source: Pahle et al. (2019).

Net Change in Emissions Induced by German Coal Phase-Out

–164

© ifo Institute

Figure 3

CONTENT

46 CESifo Forum 3 / 2021 May Volume 22

REFORM MODEL

set forth the conditions under which a European car-bon bank would be advantageous: (1) the emissions reduction pathway or carbon budget must be well defined by legitimate democratic institutions, (2) the performance criteria for implementing policy instru-ments must be well understood, (3) the carbon bank must act independently of any revenue objectives and remain unaffected by any lobbying by firms or na-tional governments, and (4) to avoid pursuing multiple objectives, the carbon bank must not be mandated to consider distributional effects between member states or in the income distribution. Nevertheless, it remains debatable whether such a bank would be fea-sible, given the problems of delegation and agreement on the optimal rules. But whether administered by a regulator or through a central carbon bank, optimal rules will be crucial for moving the EU-ETS off the slip-pery slope of discretionary intervention. Given what is at stake, environmental economists should make further research on this topic a top priority.

REFERENCES Aldy, J. E. (2020), “Carbon Tax Review and Updating: Institutionalizing an Act-Learn-Act Approach to U.S. Climate Policy", Review of Environ­mental Economics and Policy 14(1), 76–94.

Borenstein, S. (2016), “Fixing a Major Flaw in Cap-and-Trade", UC Berkeley Energy Institue Blog, https://energyathaas.wordpress.com/2016/08/15/fixing-a-major-flaw-in-cap-and-trade/.

Borenstein, S., J. Bushnell, F. A. Wolak, and M. Zaragoza-Watkins (2019), “Expecting the Unexpected: Emissions Uncertainty and Environmental Market Design", American Economic Review 109(11), 3953–77.

Carlson, A. and D. Burtraw, eds. (2019), Lessons from the Clean Air Act, Cambridge University Press.

Edenhofer, O., C. Flachsland, M. Kalkuhl, B. Knopf, and M. Pahle (2019), “Optionen für eine CO2-Preisreform. MCC-PIK-Expertise für den Sachver-ständigenrat zur Begutachtung der gesamtwirtschaftlichen Entwicklung", https://www.mcc-berlin.net/fileadmin/data/B2.3_Publications/Working-Paper/2019_MCC_Optionen_für_eine_CO2-Preisreform_final.pdf.

Edenhofer, O., M. Franks, and M. Kalkuhl (2021), Pigou in the 21st Cen­tury: A Tribute on the Occasion of the 100th Anniversary of the Publica­tion of The Economics of Welfare. International Tax and Public Finance, Springer US.

Edenhofer, O., M. Kosch, M. Pahle, and G. Zachmann (2021), “A Whole-Economy Carbon Price for Europe and How to Get There", Pol-icy Contribution 06/2021, https://www.bruegel.org/wp-content/up-loads/2021/03/PC-06-2021-090321.pdf.

Edenhofer, O. and M. Pahle (2019), “The German Coal Phase Out: Buying out Polluters, Not (yet) Buying into Carbon Pricing", EAERE Magazin.

Ellerman, A. D., C. Marcantonini, and A. Zaklan (2016), “The European Union Emissions Trading System: Ten Years and Counting", Review of Environmental Economics and Policy 10(1), 89–107.

Friedrich, M., S. Fries, M. Pahle, and O. Edenhofer (2020), “Rules vs. Discretion in Cap-and-Trade Programs: Evidence from the EU Emission Trading System", CESifo Working Paper, München.

Gerlagh, R., R. J. R. K. Heijmans, and K. E. Rosendahl (forthcoming), “An Endogenous Emission Cap Produces a Green Paradox", Economic Policy.

Harvey, D. I., S. J. Leybourne, R.Sollis, and A.M. R. Taylor (2016), “Tests for Explosive Financial Bubbles in the Presence of Non-Stationary Vola-tility", Journal of Empirical Finance 38, 548–74.

Hasegawa, M., and S. Salant (2014), “Cap-and-Trade Programs under Delayed Compliance: Consequences of Interim Injections of Permits", Journal of Public Economics 119, 24–34.

Hepburn, C. (2006), “Regulation by Prices, Quantities, or Both: A Review of Instrument Choice", Oxford Review of Economic Policy 22 (2), 226–47.

Knodt, M., M. Pahle, N. Aus dem Moore, O. Edenhofer, U. Fahl, B. Gor-lach, M. Kosch, F. Pause, G. Perino, and S. Schlacke (2020), “Wegmarken für das EU-Klimaziel 2030. Versteckte Risiken und Chancen der Sze-narien der EU-Kommission für den Pfad zur Klimaneutralität", Ariadne Kurzdossier.

Kydland, F. E. and E. C. Prescott (1977), “Rules Rather than Discretion: The Inconsistency of Optimal Plans”, Journal of Political Economy 85(3), 473–91.

Long, J. B. De, A. Shleifer, L. H. Summers, and R. J. Waldmann (1990), “Noise Trader Risk in Financial Markets”, Journal of Political Economy 98(4), 703–38.

Osorio, S., O. Tietjen, M. Pahle, R. C. Pietzcker, and O. Edenhofer (2020), “Reviewing the Market Stability Reserve in Light of More Ambitious EU ETS Emission Targets", Econstor.

Pahle, M. (2020), Schriftliche Stellungnahme zum Thema „Ökologische Aspekte des Kohleausstiegs", Internet-Öffentliches Fachgespräch des Umweltausschusses des Bundestags am 15. Juni, https://www.bundestag.de/resource/blob/700506/685f63fafc6a1056d5d-18a0747773c32/19-16-352-C_PIK-data.pdf.

Pahle, M., O. Edenhofer, R. Pietzcker, O. Tietjen, S. Osorio, and C. Flachsland (2019), “Die unterschätzten Risiken des Kohleausstiegs", Energiewirtschaftliche Tagesfragen.

Perino, G. (2018), “New EU ETS Phase 4 Rules Temporarily Puncture Waterbed", Nature Climate Change 8(4), 262–64.

Perino, G., M. Pahle, F. Pause, S. Quemin, H. Scheuing, and M. Willner (2021), “EU ETS Stability Mechanism Need s a New Design", CEN Policy Brief, https://www.cen.uni-hamburg.de/research/policy-briefs/bilder-docs/msr-policy-brief-eu-ets.pdf.

Salant, S. W. (2016), “What Ails the European Union’s Emissions Trad-ing System?", Journal of Environmental Economics and Management 80, 6–19.

Salant, S. W. and D. W. Henderson (1978), “Market Anticipations of Gov-ernment Policies and the Price of Gold", Journal of Political Economy 86(4), 627–48.

Wettestad, J. and T. Jevnaker (2019), “Smokescreen Politics? Ratchet-ing Up EU Emissions Trading in 2017”, Review of Policy Research 36(5), 635–59.

CONTENT

47CESifo Forum 3 / 2021 May Volume 22

DICE DATA ANALYSIS

The Covid-19 crisis represents one of the most se- vere global crises in modern times. The scale of the pandemic has created substantial economic turmoil, leading to a worldwide economic downturn and a sig- nificant reshaping of the global economy.

While the health consequences of the virus itself play an important role in explaining the economic effects of the Covid-19 pandemic, government re-actions to the pandemic have also played an equally, if not more, important role. To a large extent, governments did not have time to develop compre-hensive strategies, but instead had to react quickly to contain the spread of the virus. Thus, govern-ments implemented a variety of measures with- out much international coordination, resulting in enormous variation in responses to the pan- demic.

The following article presents a new tool that enhances users’ understanding of this variation based on a joint initiative to visualize the data collected by the CoronaNet Research Project, an endeavor to gather information on policies related to the Covid-19 crisis, using ifo’s DICE website. CoronaNet tracks the implementation, adaptation and, where appropri-ate, repeal of Covid-19 related policies worldwide on a daily basis. While collecting data on government responses to Covid-19 is crucial for understanding how the pandemic has affected economic outcomes, data visualization tools also play an important role when it comes to making sense of the data collected. In what follows, we provide a brief overview of the CoronaNet Research Project and the data it collects, explain how this data is visualized in the Database for Institutional Comparisons of Economies (DICE), and outline the methodology behind the CoronaNet Research Project.

OVERVIEW OF THE CORONANET RESEARCH PROJECT

The CoronaNet Research Project1 is an ongoing ini-tiative which systematically documents government policies made in response to Covid-19. To date, it has gathered nearly 70,000 policies worldwide, making it the largest database of Covid-19 government policies in the world. CoronaNet employs a fine-grained tax-onomy in order to document data, categorizing each policy into 20 broad types and over 100 sub-types. It also collects information about initiators, human and geographic targets, compliance, enforcers and timings of the policies.

CoronaNet is an international collaboration of more than 600 volunteer research assistants man-aged and organized by principal and co-princip-al investigators from all over the world. It has received internal financial support from the Hochschule für Politik, Technical University of Munich, and NYU Abu Dhabi. It has also received funding from the Peace Research Institute Frankfurt, a member of the Leibniz association.2 Its most significant external collabora-tion to date has been its participation in PERISCOPE,

1 The project website is available at https://www.coronanet-project.org/.2 Leibniz-Institut Hessische Stiftung Friedens- und Konflikt-forschung or HSFK.

Cindy Cheng*, Luca Messerschmidt*, Svanhildur Thorvaldsdottir*, Clara Albrecht**, Christa Hainz**, Tanja Stitteneder**, Joan Barceló***, Vanja Grujicx, Allison Spencer Hartnettxx, Robert Kubinec***, Timothy Modelxxx, Caress Schenk‡

Tracking Government Responses To Covid-19: The CoronaNet Research Project

* TU Munich** ifo Institute*** NYU Abu Dhabix University of Brasiliaxx University of Southern Californiaxxx Fors Marsh Group‡ Nazarbayev University

Governments around the world have taken a significant number and variety of actions in response to the Covid-19 pandemic. To understand this flood of government ac-tions, policymakers and researchers need access not only to high-quality, up-to-date data on government responses, but also tools to help them make sense of that data. In a joint in-itiative, the data collected in CoronaNet are visualized on the ifo Institute’s DICE website. The following article intro-duces the CoronaNet research project and explains some of the data collected and how they are presented on DICE.

ABSTRACT

CONTENT

48 CESifo Forum 3 / 2021 May Volume 22

DICE DATA ANALYSIS

an academic consortium of 32 EU universities investi-gating the behavioral and socio-economic impacts of Covid-19. PERISCOPE is funded by Horizon 20203 and will ensure that CoronaNet’s data collection effort can continue for three more years.

THE DATASET

The dataset provides comprehensive and granular documentation of government policies across the world for more than 200 countries. It covers the fol-lowing policy types:

1. Declaration of Emergency2. Quarantine3. Lockdown4. External Border Restrictions5. Internal Border Restrictions6. Restriction of Mass Gatherings 7. Social Distancing8. Curfew 9. Closure and Regulation of Schools10. Restrictions and Regulation of Government Ser-

vices 11. Restriction and Regulation of Businesses12. Health Monitoring 13. Health Testing14. Health Resources15. Hygiene 16. Public Awareness Measures17. Anti-Disinformation Measures18. New Task Force, Bureau or Administrative Con-

figuration19. Covid-19 Vaccines20. Other Policy not Listed Above

National-level data is available for most countries, and subnational (ISO administrative level 2) data is also collected for the following countries: Brazil, China, Canada, France, Germany, India, Italy, Japan, Nigeria, Russia, Spain, Switzerland and the United States.

The dataset not only includes information about which governments are responding to Covid-19, but it also covers who the policies are targeted at (e.g., other countries), how they are doing it (e.g., imposing travel restrictions, banning exports of masks) and when they are doing it. More specifi-cally, the CoronaNet dataset collects information on government policy actions taken in response to Covid-19 across the following dimensions (Cheng et al. 2020):

‒ The type of government policy implemented (e.g., quarantine or closure of schools).

3 PERISCOPE is funded by the European Commission under the Ho-rizon 2020 Research and Innovation program (Agreement No. 101016233).

‒ The level of government initiating the action (e.g., national or provincial).

‒ The geographical target of the policy action, where applicable (e.g., national, provincial or municipal).

‒ The human or material target of the policy action, where applicable (e.g., travelers or masks).

‒ The directionality of the policy action, where ap-plicable (e.g., inbound, outbound or both).

‒ The mechanism of travel that the policy action targets, where applicable (e.g., flights or trains).

‒ The enforcement of the policy action (e.g., man-datory or voluntary).

‒ The enforcer of the policy action (e.g., national government or military).

‒ The timing of the policy action (e.g., date an-nounced, and date implemented).

CORONANET DATA IN DICE

In a joint initiative between the researchers of the Technical University of Munich and the ifo Institute, the data collected in CoronaNet is presented on ifo’s DICE website in a user-friendly way.4 In addition to the 20 broad policy types, DICE will include informa-tion on sub-measures, when available, and present an even more fine-grained picture of governments’ policy responses.

Figure 1 shows some of the broad key policy measures that were taken on a national level in seven selected countries (Austria, the Netherlands, Germany, Italy, Sweden, the UK, and the US).5 The data refers to the proportion of a month, i.e., the number of days in a month on which each measure was in place divided by the number of days of that month. It ranges from 0 (measure did not apply) to 1 (measure applied over the whole month). When looking at lockdowns, for instance, data shows that some countries enforced them at a very early stage (e.g., Austria and Italy), while others implemented them much later (e.g., the UK) or more hesitantly (e.g., the US). All countries shown implemented quar-antining as a national policy response to Covid-19, some (e.g., Germany, the UK and Italy) earlier than others (e.g., Austria). In addition, all countries re-stricted mass gatherings early on in the pandemic. Interestingly, Italy and Austria briefly lifted these restrictions in the summer of 2020, before subse-quently reintroducing them. As early as January 2020, the first countries, such as Italy, the US, and the UK placed restrictions on their external borders, and other countries quickly followed. These restric-tions remained in place throughout the period shown (including March 2021) and are only now gradually being lifted.

4 Data and visualizations on policy types and subtypes at the na-tional level will be available at https://dice.ifo.de.5 Note that the data presented here covers national regulation and there may be additional policies at the sub-national level.

CONTENT

49CESifo Forum 3 / 2021 May Volume 22

DICE DATA ANALYSIS

DATA COLLECTION METHODOLOGY

As researchers learn more about the various health, economic, and social effects of the Covid-19 pan-demic, it is crucial that they have the greatest possi-ble access to data that is reliable, valid, and timely. CoronaNet has adopted a data collection methodol-ogy which it believes optimizes over all three of these constraints. In the following, we present an overview of CoronaNet’s data collection methodology.6

To collect the data, CoronaNet has organized more than 600 research assistants from colleges and universities around the world, representing 20 of the 24 time zones, who are actively collecting data at any given point in time. In order to join the project, re-search assistants must view a series of training vid-eos and undergo a training assessment. Once on the project, RAs communicate with regional and country managers, who oversee their work, and with the prin-cipal investigators and other research assistants, who give feedback and support.

Each research assistant is responsible for tracking government policy actions for at least one country. Research assistants are allocated depending on their background, language skills, and expressed interest in certain countries. Depending on the level of policy co-ordination at the national level, certain countries are assigned multiple research assistants, for instance, Germany or France. This data collection strategy lev-erages the benefits of widespread recruitment and

6 For more information, please see Cheng et al. (2020).

the diverse pool of country-specific knowledge across the globe.

Research assistants obtain assistance in identify-ing raw policies to code through CoronaNet’s cooper-ation with both Jataware and Overton, both of whom use machine-learning algorithms to find news reports (Jataware) and government policies (Overton) related to Covid-19. Research assistants can also check the following platforms to identify relevant policies:

‒ the information page on Covid-19 policies for a particular country on the US Embassy website;

‒ the Wikipedia page on a particular country’s re-sponse to the Covid-19 pandemic;

‒ the relevant government websites of a particular country (e.g., executive office, health ministry);

‒ newspaper coverage in a particular country (via e.g., LexisNexis or Factiva).

Once they have identified policies to code, research assistants are encouraged to organize the information they collect into country overviews and timelines, so that they can locate where a given policy fits into a country’s policy history. They are further aided in this by a custom-designed Shiny App, which visualizes existing CoronaNet data.

Research assistants then enter the data they have found into a customized Qualtrics survey. Survey questions are designed to systematize and streamline the documentation of a given government policy over a wide range of dimensions. The tool enables research assistants to document information about different

0.0

0.2

0.4

0.6

0.8

1.0

012020

02 03 04 05 06 07 08 09 10 11 12 012021

02 03

Restrictions of Mass Gatherings

Proportion of a month

Source: CoronaNet Research Project (as of March 2021). © ifo Institute

0.0

0.2

0.4

0.6

0.8

1.0

012020

02 03 04 05 06 07 08 09 10 11 12 012021

02 03

External Boarder Restrictions

Proportion of a month

0.0

0.2

0.4

0.6

0.8

1.0

012020

02 03 04 05 06 07 08 09 10 11 12 012021

02 03

Austria Germany Italy Netherlands Sweden UK US

Policy Responses to Covid-19 in Selected Countries

Proportion of a month

Lockdown

0.0

0.2

0.4

0.6

0.8

1.0

012020

02 03 04 05 06 07 08 09 10 11 12 012021

02 03

Quarantine

Proportion of a month

Figure 1

CONTENT

50 CESifo Forum 3 / 2021 May Volume 22

DICE DATA ANALYSIS

policy actions easily and efficiently by answering the relevant questions in the survey (Büthe et al. 2020). For example, instead of entering the country that in-itiated a policy action into a spreadsheet, research assistants answer the following question in the sur-vey: “From what country does this policy originate?” and choose from the available options given in the survey. This method of data collection has many ben-efits, especially in terms of ensuring the reliability and validity of the resulting data, such as by preventing data being entered in the wrong fields, standardizing responses, and reducing missed data by employing forced data entries.

CoronaNet further implements cleaning and val-idation processes, including internal and external cleaning and multiple coding, as well as an evalua-tion and reconciliation procedure, in which CoronaNet checks for discrepancies between the original coded data and the second and third codings.

CONCLUSION

Governments around the world have implemented a substantial number and variety of policies in reaction to the Covid-19 pandemic. In order to make sense of this flurry of government actions, policymakers and researchers not only need access to quality, up-to-date data on government responses to Covid-19, they also need tools to help them make sense of this data.

By providing interactive and user-friendly visu-alizations of data on government policies collected by the CoronaNet Research Project on the ifo DICE database, we want to improve researchers’ and pol-icymakers’ understanding of the pandemic and their analyses of whether, how, and to what degree these fast-changing policies have succeeded in mitigating the health, political, and economic impacts of the pandemic.

REFERENCES Cheng, C., J. Barcel´o, A. S. Hartnett, R. Kubinec, and L. Messerschmidt (2020), “Covid-19 Government Response Event Dataset (Coronanet v.1.0)”, Nature Human Behaviour 4(7), 756–768.

Marquardt, K. L., D. Pemstein, C. Sanhueza Petrarca, B. Seim, S. L Wil-son, M. Bernhard, M. Coppedge, and S. I. Lindberg (2017), “Experts, Coders, and Crowds: An Analysis of Substitutability”, V-Dem Working Paper 53.

Raykar, V. C., S. Yu, L. H. Zhao, A. Jerebko, C. Florin, G. H. Valadez, L. Bogoni, and L. Moy (2009), “Supervised Learning from Multiple Ex-perts: Whom to Trust When Everyone Lies a Bit”, in Proceedings of the 26th Annual International Conference on Machine Learning, 889–896.

Raykar, V. C., S. Yu, L. H. Zhao, G. H. Valadez, C. Florin, L. Bogoni, and L. Moy (2010), “Learning from Crowds”, Journal of Machine Learning Re­search 11(4), 1297−1322.

Sheng, V. S., F. Provost, and P. G. Ipeirotis (2008), “Get Another Label? Improving Data Quality and Data Mining Using Multiple, Noisy Labelers”, in Proceedings of the 14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 614–622.

Sumner, J. L., E. M. Farris, and M. R. Holman (2020), “Crowdsourcing Reliable Local Data”, Political Analysis 28(2), 244–262.

CONTENT

51CESifo Forum 3 / 2021 May Volume 22

MACRO DATA INSIGHTS

-5

0

5

10

15

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021ᵃ Annual percentage change (3-month moving average).Source: European Central Bank. © ifo Institute

Change in M3ᵃ

%

ECB reference value 4.5%

The annual growth rate of M3 decreased to 12.3% in February 2021, from 12.5% in January 2021. The three-month average of the annual growth rate of M3 over the period from December 2020 to February 2021 reached 12.4%.

-4

-2

0

220082009201020112012201320142015201620172018201920202021

Source: European Commission. © ifo Institute

Monetary Conditions Index

January 1999 = 0 (inverted scale)monetary

easing

monetarytightening

Average (1999–2020)

Between April 2010 and July 2011, the monetary conditions index had remained stable. Its rapid upward trend since August 2011 had led to the first peak in July 2012, signaling greater monetary easing. In particular, this was the result of decreasing real short-term interest rates. In May 2017 the index had reached one of the highest levels in the investigated period since 2007 and its slow downward trend was observed thereafter. A steady upward trend that had prevailed since October 2018 was abruptly halted in March 2020 with the onset of the Covid-19 crisis, and the index continued to decline in 2020. Its rapid growth in January 2021 was again followed by a continuous decline in the index in February and March 2021.

-2

0

2

4

6

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021

Yield spread Short-term Long-term

ᵃ Weighted average (GDP weights).Source: European Central Bank; calculations by the ifo Institute. © ifo Institute

Nominal Interest Rates ᵃ

%

In the three-month period from January 2021 to March 2021 short-term interest rates increased slightly: the three-month EURIBOR rate amounted to – 0.54% in March 2021, compared to – 0.55% in January 2021. The ten-year bond yields also increased from – 0.18% in January 2021 to 0.03% in March 2021, while the yield spread grew from 0.37% to 0.57% between January 2021 and March 2021.

100

200

300

400

500

600

700

800

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021

Euro STOXX 50 Dow Jones industrial German share index (DAX) British share index (FTSE 100)

Source: Deutsche Borse; Dow Jones; FTSE; STOXX. © ifo Institute

Stock Market Indices

January 1996 = 100

The global fears about the spread of the Coronavirus, oil price drops caused by an oil price war between Russia and the OPEC countries, and the possibility of a recession led to the stock market crash in March 2020, and global stocks saw a severe downturn in this month. Yet the German stock index DAX continued to grow in March 2021, averaging 14,508 points compared to 13,951 points in February 2021, while the UK FTSE-100 also increased from 6,584 to 6,714 in the same period of time. The Euro STOXX amounted to 3,813 in March 2021, up from 3,667 in February 2021. Furthermore, the Dow Jones Industrial continued to increase, averaging 32,373 points in March 2021, compared to 31,293 points in February 2021.

Statistics Update

Financial Conditions in the Euro Area

CONTENT

52 CESifo Forum 3 / 2021 May Volume 22

MACRO DATA INSIGHTS

EU Survey Results

60

70

80

90

100

110

120

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021© ifo Institute

EU27 Employment Expectations IndicatorSeasonally adjusted

2000 –2019 = 100

Source: European Commission.

In March 2021 the Employment Expectations Indicator (EEI) saw a forceful increase: + 6.1 points to 98.0 in the EU27, and + 6.8 points to 97.7 in the euro area.

65

70

75

80

85

-60

-40

-20

0

20

2018 2019 2020 2021

Capacity utilization

Source: European Commission. © ifo Institute

EU27 Capacity Utilisation and Order Books in the Manufacturing IndustrySeasonally adjusted

Balance %Assessment of order books

Managers’ assessment of order books reached – 13.7 in March 2021, compared to – 18.7 in February 2021. In January 2021 the indicator had amounted to – 20.3. Capacity utilization stood at 77.6 in the first quarter of 2021, up from 76.5 in the fourth quarter of 2020, showing the gradual improvement from the Covid-19 shock.

60

70

80

90

100

110

120

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021Source: European Commission. © ifo Institute

EU27 Economic Sentiment IndicatorSeasonally adjusted

2000–2019 = 100

In March 2021 the Economic Sentiment Indicator (ESI) improved sharply in both the EU27 (+ 6.9 points to 100.0) and the euro area (+ 7.6 points to 101.0), compared to February 2021. In the EU27, the ESI’s increase in March 2021 was driven by improving confidence in all surveyed business sectors (i.e., industry, services, retail trade, construction) and among consumers.

-40

-30

-20

-10

0

10

2018 2019 2020 2021

Industrial confidence Consumer confidence

Source: European Commision. © ifo Institute

EU27 Industrial and Consumer Confidence IndicatorsPercentage balance, seasonally adjusted

Balance

* The industrial confidence indicator is an average of responses (balances) to the ques-tions on production expectations, order-books and stocks (the latter with inverted sign).** New consumer confidence indicators, calculated as an arithmetic average of the fol-lowing questions: financial and general economic situation (over the next 12 months), unemployment expectations (over the next 12 months) and savings (over the next 12 months). Seasonally adjusted data.

In March 2021, the industrial confidence indicator increased by 4.6 in the EU27 and by 5.1 in the euro area (EA19). The consumer confidence indicator also in-creased by 3.6 in the EU27 and by 4.0 in the EA19 in March 2021.

CONTENT

53CESifo Forum 3 / 2021 May Volume 22

MACRO DATA INSIGHTS

Euro Area Indicators

5

7

9

11

13

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021

EA19 EU27

Source: Eurostat. © ifo Institute

Unemployment Rate

%

Euro area unemployment (seasonally adjusted) amounted to 8.3% in February 2021, stable compared to January 2021. EU27 unemployment rate was 7.5% in February 2021, also stable compared to January 2021. In February 2021 the lowest unemployment rate was recorded in Poland (3.1%), Czechia (3.2%) and the Netherlands (3.6%), while the rate was highest in Spain (16.1%) and Greece (15.8%).

-1

0

1

2

3

4

5

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021

Core inflationᵃ Total

ᵃ Total excl. energy and unprocessed food.Source: Eurostat. © ifo Institute

Euro Area Inflation Rate (HICP)

Change over previous year in %

Euro area annual inflation (HICP) amounted to 1.3% in March 2021, up from 0.9% in February 2021. Year-on-year EA19 core inflation (excluding energy and unprocessed foods) amounted to 1.0% in March 2021, down from 1.2 in February 2021.

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021-16

-12

-8

-4

0

4

EA19

EU27

Source: Eurostat.

Gross Domestic Product in Constant 2015 Prices

© ifo Institute

Change over previous year in %

According to the Eurostat estimates, GDP decreased by 0.7% in the euro area (EA19), and by 0.5% in the EU27 during the fourth quarter of 2020, compared to the previous quarter. These declines follow a strong rebound in the third quarter of 2020 (+ 12.5% in the EA19 and + 11.6% in the EU27) and the sharp decreases in the second quarter of 2020 (– 11.6% in the EA19 and – 11.2% in the EU27). Compared to the fourth quarter of 2019, i.e., year over year, seasonally adjusted GDP decreased by 4.9% in the EA19 and by 4.6% in the EU27 in the fourth quarter of 2020.

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021

Exchange rate Purchasing power parity

Source: European Central Bank; OECD; calculations by the ifo Institute. © ifo Institute

Exchange Rate of the Euro and Purchasing Power Parity

USD per EUR1.6

1.0

1.1

1.3

1.2

1.5

1.4

The exchange rate of the euro against the US dollar averaged approximately 1.21 $/€ between January 2021 and March 2021. (In December 2020 the rate had amounted to around 1.22 $/€.)

CONTENT

2021Vol. 20

Chapter 1:

MACROECONOMIC CONDITIONS AND OUTLOOK

During the first wave of the coronavirus crisis, countries im-posed strict lockdowns to reduce Covid-19 infections. This led to the sharpest reduction in GDP (almost 10 percent) since the Second World War, but the economy recovered quickly during the summer months. When the second wave hit in autumn, harsh lockdown measures were postponed in order to prevent another sharp downturn in value ad-ded. However, Covid-19 death rates reached higher levels than during the first wave. Yet, overall economic producti-on remained 2 percent under the pre-crisis level in the third quarter. The situation is expected to improve slowly until spring and economic forecasts are therefore associated with great uncertainty.

Chapter 2:

DISTRIBUTIONAL CONFLICTS AND SOCIAL CAPITAL

The ongoing coronavirus crisis and the resulting contain-ment measures have different consequences for different groups of society and across countries and thus affect so-cial capital and cohesion. Costs and benefits of lockdown restrictions and economic policy interventions are not equally shared and consequently might lead to the erosion of social capital within the European Union. In this chapter, we present policy actions suitable to counteract intergene-rational conflicts and to promote a more equal burden sha-ring, including reforms of tax and pension systems as well as the formation of more resilient institutions.

Chapter 3:

LIFECYCLES AND EDUCATION: THE CORONAVIRUS CRISIS ACROSS GENERATIONS

Education is a key investment in the future since it strongly contributes to securing intergenerational and social equi-ty. In the wake of the coronavirus crisis, schools have been closed, which will consequently lead to future income los-ses for those affected, especially already underprivileged students. Thus, school closures bear the risk of further exacerbating education inequalities. To counteract this effect, rethinking the provision of education is crucial. Hence, new ways of teaching and learning should be adopted.

Chapter 4:

BUSINESS INVESTMENT

Business investment is key for long-term economic growth and productivity. In this chapter, we discuss possible thre-ats to a successful economic recovery. Due to limited access to financing and limited reserves, small- and me-dium-sized companies may struggle to survive after the cri-sis. Public support to companies bears the risk of keeping “zombie firms” alive that would otherwise not be viable. European coordination of investment programs is needed since not all EU member states have the financial resour-ces to support desirable private investment and harmful subsidy races of individual national champions need to be circumvented.

Beyond the Coronavirus Crisis: Investing for a Viable Future

Online available here: www.cesifo.org/en/eeag-report

CONTENT

Online available here: https://www.ifo.de/en/viewpoint

ifoVIEWPOINTS

Comments by ifo President Clemens Fuest on Current EconomicPolicy Issues

CONTENT