Socially responsible agencies*
Jean Tirole Toulouse School of Economics (TSE) and Institute for Advanced Study in Toulouse (IAST), France

Search for other papers by Jean Tirole in
Current site
Google Scholar
Full access

Should agencies confine their role to their main duty or should they embrace new and desirable societal objectives? This article first discusses two emblematic examples of mission expansion: socially responsible competition authorities and green central banks. It then sheds light on the ongoing debate using contributions to bureaucratic design in economics and political science. On that basis, it warns against the risks associated with the loss of accountability, institutional conflicts, and the lack of policy coordination.


Should agencies confine their role to their main duty or should they embrace new and desirable societal objectives? This article first discusses two emblematic examples of mission expansion: socially responsible competition authorities and green central banks. It then sheds light on the ongoing debate using contributions to bureaucratic design in economics and political science. On that basis, it warns against the risks associated with the loss of accountability, institutional conflicts, and the lack of policy coordination.

1 Introduction

There have lately been serious calls for enlarging agencies’ missions to include climate change, inequality and other societal considerations. The two most prominent cases in point are the suggestions that central banks be green and competition authorities be socially responsible. A few years back this evolution would have seemed unlikely, especially for independent authorities,1 as independent-authority bashing was then the new game in town.2 Now the talk of the town is to enlarge agencies’ remits.

How did this happen? Does the shift reflect a reversal of public opinion, in particular regarding the role of banking and competition in society? Is it supported by agencies eager to appear socially responsible and thereby protect themselves against political interference? Do agencies engage in mission creep to build institutional ‘empires’? Or, on the contrary, are independent agencies using their autonomy to correct government failures? Is the evolution a welcome opening to solving some of our key – and in the case of climate change and inequality, existentialist – societal problems?

I certainly do not have an answer to all of these questions, but this article attempts to shed light on the relevant trade-offs. It is organized as follows. Section 2 discusses the two most prominent examples: the greening of central banks and the push for a public interest standard for antitrust agencies. Section 3 reports on what political and economic sciences have to say about mission assignment in the public sector. Section 4 concludes.

2 Independent agencies and the political terrain: two illustrations

2.1 Green central banks

The mission of central banks has traditionally been to pursue an inflation target and to secure financial stability. A succession of economic hardships – the financial crisis, the Eurozone crisis and the COVID-19 pandemic – led them to explore new territories however, including the fiscal one. While central banks’ policies went into the right direction in my view, ‘non-conventional monetary policy’ and the central banks’ support to countries and markets came with a few hazards too: for example the ECB’s attempt to unplug Greece in 2015 led the institution into highly political terrain. Had Greece taken a less reasonable stance and left the Eurozone, the resulting misery that would have afflicted ordinary Greek citizens could have led to the termination of ECB’s independence. Today, one may wonder about the possibility of a popular backlash against inflation and central banks’ disengagement from their narrow mission.3

Many call for central banks to take on climate change responsibilities. Integrating climate risk into stress tests is a no-brainer in my opinion, and is already in the remit of the central bank. But what about the proposal to relax prudential standards for climate-friendly bank lending? Is it necessary to run the risk of a new subprime crisis (incidentally, raising capital requirements for brown finance would already be preferable), all the more that better instruments could be employed to tackle climate change? Similarly, should the ECB purchase green bonds in order to reduce their spreads? The issue in recent years has not been a shortage of money looking for ‘parking spaces’; indeed, real interest rates have been, and still are, very low. The real issue is the lack of a business model for green investors in an environment in which authorities are loath to price carbon properly.

The concept of a ‘green central bank’ raises three issues. The first is the central bank’s lack of legitimacy. The need for a democratic mandate for the central bank to enlarge its mission is a proposal on which proponents and opponents of green central banks can come together. Yet, there is something bizarre when the polity refuses to impose a carbon price and more generally to effectively fight climate change, and yet asks the central bank (or any other institution for that matter) to do so.

The second issue is the central bank’s lack of capability. A central bank should be endowed with sufficient staff to verify emissions claims and with a methodology to define what is green. It should also be able to run cost-benefit analyses (as we want effective climate policies)? To be fair, the democratic process is often oblivious to approaches aiming at obtaining the most bang (emissions reduction) for their buck, but at least budgets are subject to a democratic discussion while a central bank’s spending is not.

The third issue is that of allocation of responsibility within government. We will return to this point, but we note right away that the absence of identification of who is really in charge of environmental policy is a concern.

To be clear: our collective procrastination has made climate change a truly existential problem and we are at a stage where every effort is worth considering. The view that the central bank can act as a ‘climate- change fighter of last resort’, while making a conviction judgment on the functioning of our democracies, is respectable. But it should indeed be a very last resort, as this organization of public policymaking runs, as I will argue, counter to our wisdom in the matter.4

a Agencies under political control

As noted earlier, mission enlargement also affects agencies under political control. A recent law in France alters the Public Procurement Code to make the integration of environmental clauses in all public procurement contracts mandatory, rather than optional. As discussed in more detail in the Blanchard-Tirole (2021) report, this well-meaning policy creates several issues, some of which are reminiscent of the green central banks debate: policy coherency (the implicit carbon price may vary enormously across jurisdictions), lack of capability (the territorial governments missing the ability to verify the bidders’ environmental claims) and electioneering-induced protectionism by municipalities and regional governments.

2.2 Socially responsible competition authorities

The proper objective for competition authorities has always been a topic of debate. An older dimension of this debate concerns whether authorities should embrace a welfare standard (consumer surplus plus industry profit) or a much narrower consumer standard, in which the profit dimension is ignored. The terms of this debate are known: profits benefit consumers if they are reinvested to produce improved products or if they serve as an incentive for the very existence of the firm and the introduction of useful services. While abuses of dominant position and cartelization can be opposed even under the more lenient welfare standard, the latter is less convincing if profits reflect wasteful expenditures (advertising, lobbying and rent-seeking etc). In the end, however, two key reasons for discounting profits relative to consumer surplus are a redistributive motive and the desire to create a simple mission.5

The idea that firms’ owners are well-off relative to the population6 was at the core of Brandeis’ doctrine. In that, there is some convergence between the advocates of the consumer welfare standard and members of the New Brandeis movement. In pointing at key competition policy issues, the two also have much in common: for example, that monopoly power is ripe with potential for abuse, including in the political domain; that dominant tech platforms raise concerns about contestability and about fairness (the levy of large fees paid by advertisers, content providers and merchants, the possibility of self-preferencing by platforms); and that preemptive/defensive mergers should be better monitored by authorities.

However, the two conceptions of antitrust diverge. Brandeisians call for adopting a ‘public interest’ standard considering societal goals such as redistribution, environment, unemployment and wage growth, political power of dominant firms and strategic independence; in antitrust parlance, the antitrust enforcer should account for ‘out of market effects’. That is, the New Brandeis movement adds stakeholders. Consumer-welfare-standard advocates retort that, while Brandeisian ‘hipster antitrust’ envisions fully legitimate goals, these goals (a) are definitely in the political domain, and (b) should be addressed using the right instruments. They also are concerned about a possible unintended consequence of the Brandeisian push for wider objectives, that the latter might favour businesses by letting them claim inflated social benefits of their M&As or their cartel agreements.

As Schinkel7 and Schinkel-Treuren8 point out, the ramifications of the stakeholder view of antitrust are many. Under limited staff resources,9 prioritizing cases according to their social impact (‘sue milk cartels before caviar ones’), ceteris paribus, should not prove controversial. Protecting nonprofits or fair-trade importers against competition by for-profits or low-cost producers, allowing cartels in the pursuit of social goals and striking deals with an oligopoly allowing the latter to collude on the rich-consumers segment against a promise of below-cost price for poorer households are all well-meaning, but may raise eyebrows for other reasons. Likewise, exempting sustainability agreements from the cartel prohibition might invite cartel greenwashing.

The EU Draft of the Horizontal Guidelines10 is well aware of the potential abuses of a stakeholder approach and tries to draw potential red lines. Indeed, its Chapter 9 has a number of interesting illustrations of what may or may not be allowed.11

a Should oligopolists be allowed to collude on ethically related dimensions?

One can think of cases in which such agreements among competitors would improve moral outcomes, given political dormancy. It is certainly difficult for industrialists to reduce their CO2 footprint and for farmers to treat animals better if their competitors do not. However, horizontal block exemption regulations (Article 101(3) TFEU in Europe) are already available to allow certain research and development and specialization agreements that can be considered more beneficial than harmful. Furthermore, as Schinkel and Spiegel12 point out, sustainability agreements may eventually hurt sustainability; Schinkel and Spiegel’s idea is that sustainability can be a dimension of competition if consumers are socially responsible, and so suppressing this dimension of competition may have the unintended consequence of reducing sustainability.

The point I want to make here is slightly different: Industry agreements along ethical dimensions should not be protected by a safe harbour and do not preclude the need for regulation of these ethical dimensions. The first rationale for this view is straightforward: such agreements create a risk of inclusion of ancillary restraints and might be a cover for price collusion. ‘Consuming less energy by reducing total output’ is a transparent excuse for brute cartelization, but some other anticompetitive, ethically motivated agreements may be hard to figure out. We also know that whatever their nature, industry-sponsored standards can be used to erect barriers to entry. Finally, self-regulation may not be reliable when it comes to checking on loopholes and enforcing industry standards. In the end, regulation will be needed.

To conclude, most would agree on the following: ideally regulation should come from the governments, not through private initiative. The absence of sufficient governmental action (as is unfortunately the case for climate change) confronts us with a second-best situation. Sustainability agreements such as a boycott of highly polluting input suppliers or a coordination on environmentally friendly packaging have the potential of raising social welfare when stand-alone responsible behaviour runs the risk of loss of market share. The competition authority may be able to acquiesce to such agreements provided it has a clear objective trading off harm to the consumer or competition against the environmental benefit (which requires putting a price on carbon13) and that it carefully monitors the industry’s compliance with its societal pledges.

b Right outcome, wrong process

The EC 2021 case against German diesel automakers for non-compliance with NOx emissions regulation nicely illustrates the authorities’ dilemma between mission and societal concerns. The non-compliers in the dieselgate were fined €2.7 billion. Interestingly, the case was prosecuted as a collusion case. It was indeed a clear-cut case of collusion. Not a standard cartel in the output market (prices, market shares) or in the input market (monopsonization, suppression of R&D competition), but rather a cartel colluding on the non-compliance with environmental norms.14 Ale-Chilet et al’s empirical work shows that there was no overregulation, and that the non-compliance was definitely welfare reducing. Case closed? Not quite. Ale et al also demonstrate that the cartel benefited car buyers!15 Interestingly the outcome was the right one, but the process for reaching it was bizarre. The EU environmental standards should have been enforced politically by ministries in charge of the environment, not by an agency in charge of protecting consumers who benefited from the cartelization. In this second-best situation, the competition authority substituted itself for failing environmental regulation.16

3 Optimal response to twin market and political failure: broad mandates or narrow missions?

3.1 Accountability

Government officials, like all individuals, are subject to moral hazard, broadly construed. They may exert insufficient effort, not (re)invest in their human capital, or neglect key issues while being obsessed with trivia. They may make poor appointments for a variety of reasons (politics, lack of information, esprit de corps, etc). They may maximize individual employees’ exit options into the private sector. They may be influenced by lobbies, campaign donations or public opinion. I am not pointing the finger at officials here, because fortunately not all behave in this way, and because similar behaviours can be observed outside the government. My point is that there must be countervailing forces making officials accountable. Making officials accountable is no simple matter, however. It depends on how performance can be assessed.

This is actually an old theme in political science. J.Q. Wilson’s celebrated book on bureaucracy17 looks at this very issue. Wilson first makes a few simple observations about the comparison between officials in government agencies and those employed by the private sector. The former face very limited financial incentives and much of their behaviour, besides intrinsic motivation, is motivated by career/image/legacy concerns. Wilson’s second observation is that government officials often are given a multiplicity of objectives. The third is that they enjoy limited autonomy. The restricted autonomy relates to the poor measurability of their performance. The incentives of a manager in charge of a specific, measurable objective, such as profit, can be aligned with this objective. This applies less in the public sector.

Wilson’s view is that it is necessary to develop a sense of mission for the government agency, eschew vague objectives and define a set of critical tasks or operational goals. A clear mission (according to him) creates an agency culture and facilitates the hiring of specialists and narrow experts. Relatedly, Wilson argues that agencies should mostly resist new tasks that differ from, let alone compete with, their core tasks.

3.2 What can go wrong with multi-factored missions?

Multi-factor missions raise two concerns.

a Too many cooks spoil the broth

For the purpose of illustration, let us return to the fight against climate change. The traditional view is that the government, through its ministry of the environment, is responsible for designing, coordinating and monitoring actions that will most effectively reduce our dependence on fossil fuels. However, the central banks, the competition authorities (federal and state/national authorities), the IMF and the multilateral organizations and the national courts etc see climate as increasingly part of their mission. Who is then in charge? How will policy coherence emerge? All being accountable may mean that no-one is.

An additional issue is that instruments are not equally efficient at achieving a given goal; some may be cheaper than others to accomplish a given target. Worried about climate change? Use carbon pricing, an ambitious green R&D programme, green public investment programmes, etc. Concerned about inequality? Use taxation, a fairer education and other standard means to create a more equitable society. It may be difficult to endow multiple agencies with the same most-efficient instrument, for exemple a carbon price for curbing CO2 emissions or an income tax for redistributing income.

The plethora of cooks finally makes it difficult for economic agents to know what is expected from them. Take a firm’s green policy choices. The firms must be knowledgeable about how a myriad of municipalities and regional governments will assess their ecological performance when tendering, how the central bank and sovereign wealth funds will ESG-rate their bonds when issuing them and how DG Comp, national competition authorities and various courts will account for societal objectives in their rulings when contemplating an industry agreement or a unilateral behaviour. Legal uncertainty is magnified by the lack of precise definition of how to measure and weigh societal goals, and the multiplicity of agencies in charge increases the bureaucratic costs involved in trying to second guess their future decisions.

b Functioning of agency

Multiple missions given to a single agency may reduce accountability. Most obviously so when tasks are incompatible: one cannot ask an agency to act contrarily, as when for example one asks an electricity utility to both sell more electricity and advise consumers on how to conserve energy.18

Multiple missions also lead to undefined expectations as to what is expected from the agency. Because the public, the courts, or the parliament must be able to hold agencies accountable, one would ideally define how one will measure the agency’s contribution to each of its goals and weigh the various dimensions of performance.

Creating fuzzy missions for agencies that so far have performed decently, may not be a good move. Furthermore, there is a danger that governments would be encouraged, even more than now, not to assume their responsibility, as they can pass the buck to the agencies.

Creating fuzzy missions for agencies that so far have performed decently within their narrow mandates is a risky move.

Finally, there is a risk of ‘mission substitution’ if the agency’s resources are not expanded to reflect the additional missions. Given its limited resources, the agency may be forced to detract resources away from its core mission to deal with the new missions. This will be especially likely if the new missions are high on the agenda of politicians with influence on the career prospects of the agency leadership. Overall, the agency leadership may end up devoting a disproportionate amount of time and resources to the new missions to the detriment of its core objective.

3.3 The economics of missions

Even if tasks are not incompatible and expectations are clearly specified, multi-factored missions generate incentive issues. Dewatripont et al19 show that expanding the set of tasks pursued by the agency typically reduces total agency effort (‘grasp all, lose all’). Put differently, the agency faces higher incentives when given a focused mission than when allocated multiple missions.

Another issue is that multiple missions create uncertainty as to which mission the agency is pursuing. In this respect, narrow specialists have more accountability; because they are less tempted to follow other missions, one can more easily assess them on what they have been trying to achieve. This connects with Wilson’s view that mission-oriented agencies are narrow-minded but effective; and that their management is accountable, as it cannot take refuge in a correct or invented pursuit of a specific dimension of its overall mission.

A valid criticism of narrow missions is that unforeseen circumstances may arise that vindicate the agency exceeding its power. As I argued, a central bank confronted with a banking or Sovereign crisis and a political deadlock is vindicated in embracing innovative policies. On the other hand, climate change and inequality are unfortunately no longer unforeseen circumstances.

3.4 Why are some agencies independent?

Government is the quintessential multi-mission actor. In contrast, the delegation of certain specialized missions to independent agencies has allowed the pursuit of more focused missions. Independent central banks have proved better at taming inflation and regulating banks than their politically controlled predecessors. Independent competition authorities have pursued consumer interests much more consistently than was done in the past.

The expansion of missions exposes agencies to the loss of independence since independence requires limited and mandated powers.

But independence comes with duties. The agencies do not choose their mission and the broad lines of what is expected from them are clearly specified: independence requires limited and mandated powers. The expansion of missions to the political domain therefore exposes agencies to the loss of independence.

3.5 Does competition hinder ethical behaviour?

In the case of competition policy and related policies (trade liberalization, competitive procurement), there is a widespread suspicion that competition hinders moral behaviour. That is, it is argued that not only does the market fail to solve externality or distributional issues, but that it even promotes unethical behaviour. For example, the ‘replacement-logic concern’ holds that otherwise-ethically-motivated managers may not behave ethically because they know that if they did, someone else would supply a less ethical product in their place: dictators would find other ways to acquire weapons, consumers would find other doctors to prescribe opioids or sick leaves, corrupt officials would grant the contract to a less scrupulous supplier, or the race would be won by an athlete taking performance-enhancing drugs. In all these examples unethical behaviour increases demand for firms.

Mathias Dewatripont and I20 take on board this replacement-logic concern, that competition potentially puts pressure on firms to lower their ethical standards. Yet the article issues a strong warning against a sweeping condemnation of markets as immoral. First, it shows that under flexible prices, a more intense competition leaves moral choices unchanged; more competition raises the elasticity of individual firms’ demand; the higher potential market-share gain of unethical assertiveness is however compensated by a lower price-cost margin. It also shows that if prices are regulated (by an authority, a platform or a franchisor), competition boosts ethical behaviour if consumers are socially responsible, that is when their demand increases with a more ethical offering (as is the case of consumers demanding fair trade products or responsible investment). It is only when prices are regulated and unethical behaviour increases demand for firms that the replacement logic leads to a deterioration of ethics as competition becomes more intense.

The article shows that what is relevant for ethics is more incentives (the power of managerial incentives or the firms’ corporate form) than competition. The message is that one should not issue a wholesale condemnation of markets on ethical grounds, and thereby dispense with competition policy, anti-bottleneck regulation, and competition through trade without careful thinking about the environment in which competition takes place. And one should be wary of corporate incentives to push this ‘immoral markets narrative’ when it is actually false.

3.6 ESG for firms but not for agencies?

The State’s dysfunctionality implies that civil society is the last bulwark of social responsibility. It may seem paradoxical that one may favour corporate and individual ESG behaviour with regards to investment, consumption and work choices, and yet have reservations about agencies’ ESG approach. Indeed, civil society’s ESG behaviour is subject to some of the same reservations as for agencies. A common criticism of ESG behaviour is that it is too multidimensional and too imprecise in its measurement of various ESG components (a situation that taxonomies and ESG rating agencies are trying to remedy).21 Furthermore, and to the extent we care only about effective ESG actions (about impact, not appearances), one should make sure that these actions achieve their goal; indeed, there is a myriad of examples of actions that look green and yet do nothing to reduce our CO2 footprint. Overall, both corporate and agency greenwashing must be eschewed.

My hunch is that corporate ESG and agency ESG are different, though. In the latter case, there are fewer checks, such as some skin in the game for management or hard oversight by the board of directors. The questions of whether we trust management not to pursue their latest fad or promote the interests of their friends or favourite cause instead of the common good, and of the ability and willingness of the board to discipline management in this multi-factor context, arise in both contexts, but in varying degrees. Whether this hunch is correct is ultimately an empirical question, on which we admittedly have very little evidence.

4 Concluding note

The goals envisioned by advocates of socially responsible agencies are unquestionably worthy. And it is easy to envision examples in which a broader agency remit could lead to a better environment, less inequality, or more respect for animal rights. We should, however, be wary about mission creep.

The first reason for this cautionary note relates to the implementation of the new missions: corporations may be eager to push for policies that involve consumer or taxpayer money, but less eager to actually monitor and implement the policies when they are enacted. The second reason is that, even if the new remit is put on record, establishing the agency’s legitimacy, the agency’s incursion into the political domain may compromise its independence. Third, the agencies’ new duties may give an additional excuse to the government not to act when it is precisely its failure to act that creates the impetus for agency mission enlargement.

While the goals envisioned by advocates of socially responsible agencies are unquestionably worthy, we should be wary about mission creep.

Fourth, and crucially, mission creep may create dysfunction in the organization of public decision-making. Having many agencies in charge of a given policy may imply that no-one is really accountable, and that turf wars and policy coordination failures emerge; it may also increase the cost of doing business by forcing firms to second guess the likely policies of a myriad of public decision-makers. Overall, ‘too many cooks spoil the broth’. Furthermore, and following the adage ‘grasp all, lose all’, a fuzzy mission may create a lack of accountability at the level of the agency.

  • 1

    The calls are broader than independent agencies, though. All institutions are asked to be socially responsible, whether independent or not, private or public, as we will later note. The case of independent agencies is particularly interesting because of their status.

  • 2

    For example, many in Europe (and not only the leaders of populist movements) called for the ECB to be put under elected officials’ control, or for DG Comp to be subject to EU governments’ oversight when making decisions that touch on national interests, such as merger reviews and abuses of dominant position. Cases in point are outrages following a rise in key interest rates of the central bank or the prohibition of Siemens' proposed acquisition of Alstom.

  • 3

    Even though some of the inflation is either imported or of a budgetary nature.

  • 4

    For further discussion of the greening of central banks, see e.g. the synthesis chapter in

    Blanchard O., and Tirole J., The Major Future Economic Challenges, report of the international commission mandated by the President of the French Republic, France Stratégie (2021).

    • Search Google Scholar
    • Export Citation
  • 5

    A third reason is formalized by Armstrong and Vickers in

    Armstrong M., and Vickers J., (2010) ‘A Model of Delegated Project Choice’ (2010) 78(1) Econometrica 213244

    , who look (say) at the optimal choice of acquisition by a firm under the welfare and consumer standards. The consumer standard may prevent desirable mergers from the point of view of total welfare, but also pushes the firm to account for consumer welfare when selecting which party to acquire. See also M.P. Schinkel and L. Treuren, ‘Green Antitrust: Friendly Fire in the Fight against Climate Change’ (2020), available at: (accessed 13 January 2023).

  • 6

    How much so depends on the allocation of stockholdings. Systems with funded pensions have a more diverse population owning a stake in the financial markets. But in all systems the average shareholder and the average debtholder are much better off than the average citizen.

  • 7

    Schinkel M.P. , '‘On Distributive Justice by Antitrust: The Robin Hood Cartel’ ' (2022 ) 18 (3 ) Journal of Competition Law & Economics : 584 -612.

  • 8

    Schinkel and Treuren (fn 5).

  • 9

    Resources are already quite limited today. Peeperkorn in

    Peeperkorn L. , '‘Competition Policy is not a Stopgap’ ' (2021 ) 12 (6 ) Journal of European Competition Law & Practice : 415 -418.

    notes that they would be even scarcer (in relation to the needs) under Brandeisian antitrust. To the extent that the authority internalizes societal concerns, any industry ought to come under its scrutiny, even perfectly competitive ones. He also notes that while the discussion has focused mainly on horizontal agreements, vertical ones may also have non-negligible societal impacts and therefore fall under the new remit.

  • 10

    Communication from the Commission (2022) ‘Guidelines on the Applicability of Article 101 of the Treaty on the Functioning of the European Union to Horizontal Co-operation Agreements’ (review and proposed amendments to 2011 guidelines).

  • 11

    Among Member States, Austria, Greece and the Netherlands have cautiously moved in the direction of softening cartel enforcement for sustainability agreements. For example, the Dutch ‘Guidelines on Sustainability Agreements’ have been invoked to allow Shell and TotalEnergies to collaborate in the storage of CO2 in empty natural-gas fields in the North Sea.

  • 12

    Schinkel M.P. & Spiegel Y. , '‘Can Collusion Promote Sustainable Consumption and Production?’ ' (2017 ) 53 International Journal of Industrial Organization : 371 -398..

    • Search Google Scholar
    • Export Citation
  • 13

    Interestingly, the Dutch competition authority ruled against an agreement regarding the closure of coal burning electricity plants. The closure would have reduced electricity capacity production in the Netherlands by 10% and led to price rises in electricity and extra costs of €75 million per year for Dutch consumers. In the end there would have been no environmental benefits as the number of allowances in the European Trading System would not have changed. But in the counterfactual situation where the coal plant closure would have reduced CO2 emissions, the authority valued environmental benefits of avoided emissions at €30 million. Such computations demonstrate the need for the agency to be given a carbon price (whether the ETS one or another price). Most estimates (prior to the Ukraine invasion) pointed at a carbon price of 40 to 50 € to make coal plants unprofitable, much lower than the carbon price that would be needed for compliance with the Paris agreement.

  • 14


    Ale-Chilet J., , Chen C., , Li J., and Reynaert M., ‘Colluding against Environmental Regulation’ (2022) mimeo in which they argue that collusion helped individual automakers to produce below the environmental standard.

    • Search Google Scholar
    • Export Citation

    First, all engaging in non-compliance lower the risk of a compliant competitor reporting the violation. Second, when some comply and others do not, there is little suspicion of overregulation and fines for non-compliance may be high; when no-one complies, there is some ambiguity on producers’ ability to comply.

  • 15

    The companies for example agreed on the size of onboard Adblue tanks, leaving much more space in the trunk for consumers.

  • 16

    Another illustration is the German competition authority’s investigation of FaceBook’s data use.

  • 17

    Wilson J.Q. , Bureaucracy. What Government Agencies Do and Why They Do It , (Basic Books, 1989 ).

  • 18

    In a paper on the role of advocacy, Mathias Dewatripont and I argue that it may be better to have advocate government agencies rather than a single, multi-mission one, even though the latter can more easily arbitrate among conflicting objectives. The potential conflict among advocates of course calls for the existence of a referee, as when a prime minister decides when the ministry of the economy and a spending ministry are in conflict. See

    Dewatripont M. & Tirole J. , '‘Advocates’ ' (1999 ) 107 (1 ) Journal of Political Economy : 1 -39.

  • 19

    Dewatripont M., Jewitt I. & Tirole J. , '‘The Economics of Career Concerns. II: Applications to Missions in Organizations’ ' (1999 ) 66 (1 ) Review of Economic Studies : 199 -217.

    • Search Google Scholar
    • Export Citation
  • 20

    Dewatripont M., and Tirole J., ‘The Morality of Markets’ (2022) mimeo.

  • 21

    See, e.g., The Economist‘s special report on ESG investing (23 July 2022), which concludes with ‘Measure less, but better’.

Contributor Notes

This article is an expanded version of a November 2022 keynote speech delivered at the EU DG Comp conference on ‘Making Markets Work for People’. I am grateful to participants for helpful comments as well as Daron Acemoglu, Olivier Blanchard, Mathias Dewatripont, Cani Fernandez, Jorge Padilla, and Maarten Pieter Schinkel for useful discussions. The author is honorary chair of the Toulouse School of Economics, whose Digital Center receives financial support from sponsors, see (accessed 13 January 2023).