1.745 Germans turned to the Bundesverfassungsgericht claiming that the European Central Bank (ECB) undermined democratic control of the use of state power in Germany. They believed that the new policies of the ECB usurped economic competences that should, according to their constitution and the Treaties, stay in the hands of the national parliament where they can be influenced by the voters. The German court followed the reasoning of the plaintiffs and asked the Court of Justice of the European Union (ECJ) to review the validity of the public sector purchase program (PSPP) that was established by the Governing Council of the ECB. The ECJ recently delivered her ruling in that case. It found that there was no reason to invalidate the program.
The alleged threat to democratic control of government is only mentioned once in the judgement, namely in the description of the claims before the referring court. The ECJ nowhere considers the principle of democratic control of government (in article 2 and 10.1 TEU) for the interpretation of the powers transferred to an independent central bank. The concerns of the Bundesverfassungsgericht remain unaddressed. The ECJ chose other methods of interpretation. This blog will provide a critical review of those choices made by the Court when trying to develop and enforce a standard of review for the compliance with the substantial constraints on monetary policy.
Although those choices of the Court are understandable when viewed separately, they lead to an overall unsatisfactory outcome. They transform substantial constraints into formal constraints, make judicial review of substantial constraints ineffective and undermine the capacity of a treaty to be used as a trust building commitment device. In so far as the participation in the Union, the transfer of powers to the Union and application of EU law in the Member States is conditioned on those substantial constraints by the political or legal constitution of the Member States, the Court, that seems to be unable or unwilling to grant effective protection to those substantial constraints, undermines the full effect, uniform application and further integration.
This blog post consists of five parts. The first discusses the interpretational methods used by the Court in this case. The second and the third title will illustrate how these interpretational methods were applied in delineating and enforcing the mandate of the central bank and the prohibition on monetary financing. The fourth title summarizes the problem those interpretational methods create for European integration and the fifth will discuss one of those problems in detail. More details about the program and similar procedures can be found in my previous blog post on the Advocate General’s opinion in the same case.
1. Interpretational method
The Treaty on the Functioning of the European Union (TFEU) and statutes of the European System of Central Banks and the European Central bank (the statutes) both contain provisions with substantial constraints on the competences granted to the central bank. They limit the competences transferred to the ESCB and the use of those transferred competences by the ESCB. Several of those provisions are however not the kind of specific rules of conduct that would allow direct application without any further interpretation in each case. For this reason the judges are obliged to make choices in the application of the law. The ECJ performs this law making activity in three stages in the cases on the Economic and Monetary Union (Pringle on the ESM, Gauweiler on OMT and Weiss on the PSPP). First, it cites the provision and recalls the general goal behind the provision. Second, it translates the principles behind the provision into a specific standard. It prefers to endorse those standards that were already developed in secondary law. This lifts choices from the policy level (secondary law) to the constitutional level (primary law). In general the compliance regarding those endorsed standards dependents on a factual assessment. Third, it applies these standards. Finding itself unfit to assess the facts, it leaves the assessment to the institutions under review. The ECJ only requires that those institutions motivate why they believe that they comply with each point of the standard endorsed in the second step. The marginal control of this motivation is rather weak because the judges believe that they do not have the skills to do that.
2. Mandate of the ESCB
Step 1: The goal behind the provision
The treaty grants the exclusive competence for monetary policy to the ESCB (article 127.2 TFEU and article 3 statutes). However, the treaty and its protocols do not define the meaning of monetary policy.
The ECJ interprets this monetary competence as allowing the ESCB to use the instruments granted to the ESCB to pursue the goals set for the ESCB. With this interpretation, it creates an overlap between monetary and economic policies. The overlap comes into existence because the Member States are allowed to use the same instruments used by the ESCB to achieve their own economic goals. The ESM, for example, can – just like the ECB – provide credit to financial institutions and buy government bonds on the secondary market. The difference in the goals pursued by monetary and economic policy is not able to limit this overlap. Those goals are abstract and not self-limiting. It is also difficult to review them. The specific goals of each decision are only known to the actual decision makers (in economics: private information), in this case the voting board members in the Governing Council of the ECB. Except for God, nobody knows whether those board members voted in favor of the employment of those instruments – that are available both to the ESCB and ESM – to pursue monetary goals (for which they are competent) or economic goals (for which they are not). In so far as the judge would try to deduce this private information from the behavior of the policymaker, he will change the criterion from the goals pursued to the behavior exhibited. In that case he should be explicit about that new real behavioral standard.
An alternative would have been to interpret the monetary competence as what was regarded as the general practice in monetary policy. A practice is more concrete and self-limiting. The choice in favor of such a practice-oriented interpretation can be justified through the idea that the meaning of words is determined by their use. The policies that fall within the scope of a conventional interpretation of what monetary policy means have a more limited effect on and overlap with the economic competence of the Member States. The fact that this interpretation compares past behavior (the practice) with current behavior (a program under review) allows it to use publicly available information and would make it easier to control. A program that deviates significantly in its approach, volume, conditions and scope from what has been done in the past, was not foreseen, not assigned and not legitimated by consent of the Member States. If such a practice-oriented interpretation would have been adhered to, the unconventional monetary policy would – by definition – only have been possible after the Member States transferred this new power to the central bank through an amendment to the treaty.
In that sense, he practice-oriented approach has an interesting future in relation to the rest of the treaty. It is in full compliance with the idea that the Union only has the powers that were conferred by the Member States (article 5 TEU) in accordance with their constitution (article 54.1 TEU). Those constitutions demand that the assigned powers should be specific, meaning that it should be clear to both the one granting the power and the one receiving it which competence is actually transferred. If transferred powers are interpreted in an abstract way (based on vague goals) this would not be possible. If the powers are interpreted in a concrete way (based on practice) this would be possible. In other words, the alternative interpretation allows for an informed consent from the Member States.
Step 2: Lifting secondary law standards to primary law
The primary goal of monetary policy described in the treaty is price stability (article 127.1 TFEU). There is neither a definition of price stability nor a specific target in the treaty.
On a policy level, the Governing Council of the ECB initially interpreted price stability as a “year-on-year increase in the Harmonised Index of Consumer Prices (HICP) [not asset prices] for the Euro area [not each individual Member State] below 2%” on the medium term, being symmetrical incompatible with inflation and deflation. However, they believed that this 2% was a reference value and not a target they were trying to hit (Otmar Issing, Jackson Hole 2002). Inflation targeting was, according to the ECB’s chief economist, an epistemological impossibility. Reliable models to predict inflation, in his view, were a necessary precondition to perform inflation targeting. Those models, however, depend on an estimation of the output gap, which is notoriously difficult to estimate for the recent past. More than half of the time (!) in OECD and IMF estimates for G7 countries between 1998 and 2010, even the plus and minus sign of the estimated output gap is revised afterwards (Otmar Issing, IMF Statistical Forum 2015, 24:17-26-03).
In 2003 the Governing Council of the ECB refined its interpretation of the mandate as “below, but close to, 2%” to make it clearer that it was not ignorant of the risk of deflation, that it understood that the HICP might overestimate the real inflation and that the nominal character of the zero-lower-bound would grant it the power to set the interest rates over a wider range (namely including a real interest rate of -2%). At the time professor Otmar Issing was still the chief economist of the ECB and inflation targeting was still understood as a mirage. His successor Jürgen Stark shared this views (Jürgen Stark, 2010), and was very explicit that the ECB was not in the business of inflation targeting (Jürgen Stark, Central Banking Conference 2010). Only under the presidency of Mario Draghi has this refinement been reinterpreted as a very specific inflation target that they would try to hit, more specifically somewhere between 1,8% and 1.9%. He made that task even more difficult by focusing on core-inflation (which is lower) rather than the HICP (Lex Hoogduin, 2018). The difficulty that the ECB now experiences in achieving its own policy goal, does not mean that it is unable to achieve the goal in the treaty (price stability). The ECB only finds it difficult to achieve its own reinterpretation of its own refinement of its interpretation of that mandate (namely inflation targeting bellow but close to 2%). It is a self-inflicted problem without serious treaty relevance.
The ECJ endorses the reinterpretation of the refinement of the interpretation of the goal as the interpretation of the treaty provision, as if this interpretation was enacted in the provisions in the Treaty of Maastricht. In this act of interpretation, the Court transforms an act which should have been under treaty constraints (a policy translation of its own mandate by the Governing Council of the ECB) to the level of the constraint. This creates several problems. (1) The Court creates the risk that the ECB can now set its own competence through interpretation. It can make its own policy goals a ‘necessary’ implication of the treaty. (2) This will most certainly be the case if the ECJ is, as is currently the case, only willing to review that the interpretation of the goal is not “vitiated by a manifest error of assessment”. This allows the ECB to change that interpretation as long as it is not manifestly erroneous. (3) This risk is even more present if the ECJ performs the review of reasonability itself when the Court admits it does not have the training and expertise to do so (as we will explain in step 3).
Step 3: Dropping fact finding for a fact-dependent standard
The compliance regarding the substantial limits on the monetary policy (the treaty), understood to be a goal (step 1), for which the standard of review is reflected from a changing policy practice (step 2), is now dependent on the assessment of an economic variable (namely the estimated inflation in the medium term). If no risk for the inflation level would exist or the means used were not able to raise it, this would be a rare case in which the use of the instruments that can both achieve monetary and economic goals would be forbidden. This means that the Court ended up interpreting the treaty in a way that makes compliance or non-compliance with the treaty dependent on a factual economic variable that can only be assessed according to the methods of economic science (if it can be assessed at all; see problems discussed by Otmar Issing in step 2). Of course, the Court recognizes that it does not have the expertise to assess this variable that itself made the ultimate criterion of its standard of review.
The ECJ at that point deviates from the expected behavior of a judge. A judge who runs into the limits of his own expertise is expected to appoint a neutral expert or commission of experts to assess the facts or, if the assessment is not clear-cut (as is the case with almost all evidence), must explain how strong or weak the evidence is. The Court is, according to article 25 of the 3th protocol to the TFEU, competent to appoint such experts, but does not use this competence. The lack of expertise utilized to assess the facts is thus not an external factor, but a choice made by the Court. This choice makes it impossible for the Court to effectively asses the facts on which their judgement depends according to the standard that same Court chose to adhere to in step 2.
Taking into account their own lack of technical expertise in economics, the ECJ only demands that the reasoning of the institution under review is not “vitiated by a manifest error of assessment”. Somewhat surprising, taking into account its own stance on its own lack of knowledge in this field, the Court performs this review of manifest errors without the help of independent neutral experts. This makes it practically impossible for them to guarantee even the absolute minimum they promised, namely a judicial review for manifest errors in the assessment. If the parties in the proceedings show contradicting evidence on possible errors in the reasoning of the ECB, the ECJ will not be able to choose between them on the basis of any expertise on the subject, and will be tempted to follow the party it generally trusts, namely the institutions under review. This makes the judge dependent on the party under review.
The result of the third step is that a substantial limit professed in the treaty now ends up being a duty to argue in a not manifestly wrong way (assessed by the judges who believe they do not have the expertise to perform the assessment) on the standard reflected from secondary law that the institution under review can change. In others words, the ECJ transforms a substantial constraint into a purely formal constraint.
3. Prohibition on monetary financing
Step 1: The goal behind the provision
The provision on the prohibition on monetary financing of government debt states that “Overdraft facilities or any other type of credit facility … in favor of … Member State … shall be prohibited, as shall the purchase directly from them … of debt instruments” (article 123.1 TFEU). It adds that “Any measure, not based on prudential considerations, establishing privileged access by … governments … shall be prohibited.” (article 124 TFEU) The statutes allow the ESCB to “operate in the financial markets by buying and selling outright” (article 18.1: “Open market and credit operations”). In the last quote, the ECJ reads that “Article 18.1 … permits the ESCB, … to operate in the financial markets, inter alia, by buying and selling outright marketable instruments, which include government bonds”. The general rule (a mandate to buy on the market) precedes the specific rule (a rule regulating specific types of those instruments), it seems.
The general goal behind article 123 TFEU, according to the ECJ, is to prevent the reduction of the impetus towards sound budgetary policies. The ECJ makes this impetus the real criterion for compliance with the treaty provision. After this first step of interpretation, the provision is no longer a prohibition on the monetary financing of government debt (that is the monetarization of debt), where only the market preferences, demand and believes determine the amount of debt one can raise and the price paid for that debt (that is market discipline), but a general goal that can also be achieved by any other means preferred by the central bank.
This is exactly the same approach as the ECJ had to article 125 TFEU in the Pringle case. There it used scare-quotes around the term “no bail out clause” and explicitly expressed that the article was not a no bail out clause. Even after stating that the preparation explains that “[t]he prohibition laid down in Article 125 TFEU ensures that the Member States remain subject to the logic of the market when they enter into debt, since that ought to prompt them to maintain budgetary discipline.”, the Court ignored the specific choice (market discipline) and allowed the use of any means (other than a no bail out rule) to achieve the abstract goal behind the provision. This allowed the Member States to bail each other out as long as they took other measures to provide the bailed out Member States with an incentive to adhere to the principles of sound budgetary policies.
Both examples show the problem with the goal-oriented interpretation of the ECJ. If there is compliance with the treaty as long as the general goal behind a provision is reached, the more abstract the goal chosen by the Court, the more will be allowed. Everything can be allowed as long as the general goal is as general as is required to allow both the provision and the program under review to become specific choices that could realize the abstract goal.
An alternative would have been to look to the general practice in countries that had a prohibition on monetary financing and the explicit exclusion/inclusion of some practices by the negotiators of the treaty. This alternative interpretation would have provided for a stronger criterion than this goal-oriented approach of the ECJ. Those standards might allow the ECB to buy some government debt, but would not allow them to buy up to one third of the outstanding volume.
Step 2: Lifting secondary law standards to primary law
Council Regulation (EC) 3603/93 states that “purchases made on the secondary market must not be used to circumvent the objective of that Article.” This consideration was lifted to the level of primary law by the ECJ in the OMT-case and interpreted as meaning that the purchases on the secondary market should not have an “effect equivalent to that of a direct purchase of government bonds”. That equivalent effect would take place if “potential purchasers … knew for certain that the ESCB was going to purchase those bonds within a certain period and under conditions allowing those market operators to act, de facto, as intermediaries for the ESCB for the direct purchase ….”. In the PSPP the ECJ clarified that the required certainty on the bond bought, the period after which they would be bought and the conditions of that purchase should be absolute to violate the prohibition on monetary financing.
The ECJ in this second step of interpretation replaced the prohibition of monetary financing and privileged access with the following three demands: (1) the ESCB should buy their governments bonds on the secondary market, (2) it should not create an absolute certainty about buying a specific individual bond at a specific day against specific conditions and (3) it should be constructively ambiguous on the duration of the program or even threaten with exclusion if countries deviate from the principles of sound budgetary policy. Where the prohibition of circumvention and the preservation of budgetary discipline sounded like reasonable interpretations (although they already made abstraction of the specific choice of the treaty maker to adhere to market discipline), the level of certainty the ECB has to provide to market participants and Member States to breach the prohibition on monetary financing, is no longer a reasonable interpretation.
All three criteria are economically flawed and can neither sufficiently protect the impetus of sound budgetary policies, nor prevent equivalent effects to the purchase of government bonds on the primary market, as we discussed in more detail in a previous blog post. (1) Due to the fact that bonds on the primary and secondary markets are substitutes, the effects of purchases on market participants and Member States are equal. (2) The market players and governments act on expected values, not minimal pay outs. Certainty on the level of each bond is not required to change their behavior. (3) The ambiguity and threat are not credible. The expected effects of a sudden and significant change in the policy would go against revealed preferences of the Governing Council of the ECB.
Step 3: Dropping fact finding for a fact-dependent standard
It becomes almost impossible not to reach the threshold in the standard chosen by the ECJ. Not a single fact suggested by the ECB is openly rejected by the Court. In essence it is not the ECJ but the ECB, the party under review, which is assessing the facts in its own case. There is no serious deliberation nor any use of independent neutral experts to assess them.
Two examples might help to illustrate this point:
- The ECJ accepts that the ESCB is not circumventing the prohibition to buy on the primary market, when the ESCB buys the bonds only a few days after they were emitted. That is in relation to bonds that can have a maturity up to 30 years and 364 days.
- They believe that only countries with a healthy budget are included because of “stringent eligibility criteria based on a credit quality assessment”. These are criteria that allow the national banks to buy bonds of for example Italy.
It is not the technical and complex nature of monetary policy that makes it impossible for the Court to assess the facts that are relevant for the judicial review of the compliance with the standard to which the Court chose to adhere. It is its choice not to appoint any neutral experts.
It is not the treaty that makes the Court dependent on such a factual assessments. It is its choice to elevate policy choices and goals made under primary law to the level of primary law itself. These standards developed under primary law are however designed for a different use, namely by policymakers and experts trying to set the optimal policy, and they are therefore unfit for judicial review that tries to enforce the constraints on those policy choices.
It is not the treaty text that makes the Court depended on the review of compliance with abstract goals. It is its method of interpretation which gives meaning not through the way that words are used, but through some abstract goals behind them.
This ineffectiveness of the judicial control of substantial standards performed by the ECJ is not without relevance for European integration. Through the three steps of interpretation a set of substantial constraints were effectively transformed into purely formal constraints. In essence, the Governing Council of the ECB is now obliged to argue on the point of a standard of review that was mirrored against the ECB‘s practice and that the ECJ is not able to control.
One can argue against substantial constraints on policymakers (as do many Neo-Keynesian macro economists who believe that broader discretionary powers are needed to make our monetary union workable). However, that is not the choice made by the treaty makers. If one would like to abolish those substantial constraints, there is a procedure for that. And we should keep in mind that some countries, like for example Germany, would no longer be able to participate in the monetary union if these substantial constraints were no longer in the treaty.
Endangering the full effect, uniform interpretation of EU law and the transfer of powers
If the substantial constraints (constraints on competence and the protection of fundamental rights and liberties) are not upheld by the ECJ, the constitutional conditions for direct effect of EU law within Member States will no longer be fulfilled and constitutional courts will be willing to intervene in the most extreme cases. This decentralized review of the substantial constraints by constitutional courts, instead of centralized review by the ECJ, is however problematic from the perspective of European law because it means that the full effect and uniform interpretation are undermined.
An even more worrying effect of the interpretational technique that transformed substantial constraints into formal constraints, is that it undermines the belief that treaty provisions can be used as a commitment device to build trust and smooth the transfer of powers to the Union. The remainder of this blog post will discuss this problem in more detail.
5. Commitment device
The calculus of consent
If (potential) parties to a (potential) treaty want to assess the expected net payoff of cooperation instead of non-cooperation, they will look into the expected advantages and disadvantages of that cooperation. If the cooperation is not transactional, it might establish a common decision-making process. This common decision-making process can be governed by a rule of unanimity. In that case every party can take the cooperation captive by withholding its vote. The cost of non-decision or ransom payments to the withholding minority can be lowered by lowering the number of votes needed to decide. This however creates another cost, namely the cost of being overruled by others, that is to be forced to do something with which one disagrees. The ideal majority-rule in the case of a common decision-making procedure is that number of votes for which the expected marginal costs for undergoing the non-decision making and ransom payments equals the expected marginal cost of being overruled by the others. If the expected payoff of the cooperation under the ideal majority-rule is lower than the payoff of non-cooperation or cooperation through another means than common decision-making, the common decision-making process will not be established.
If the expected gain of a common decision-making process is negative, one can look for other remedies than the number of votes needed in the decision-making process, to lower the costs of sharing such a common decision-making process. The traditional solutions in constitutional law are: (1) limiting the competences of the decision-making body to specific topics, namely through assigned powers and explicitly excluded powers, (2) limiting the use of that decision-making power to a necessary minimum, namely through proportionality and subsidiarity, (3) limiting the type of obligations that can be created, namely regulation, directives or recommendations, (4) adopting principles that limit the room for policy-making, (5) adopting fundamental rights and liberties and (6) protecting the constitutional identities. If the parties put these solutions in the agreement for further cooperation, they are trying to lower the expected cost of common decision-making power through commitments that can be enforced by the courts. If the remedies are not effectively enforced, they will no longer be able to lower the expected cost of common decision-making and will not be able to raise the preferred level of common decision-making or the willingness to participate in such a common decision-making process.
The amendments to the treaties relating to the European Union added all of the enumerated remedies to the common decision-making process. However, the effectiveness of the judicial review has been criticized on all fronts, by politicians, legal scholars and even constitutional courts. And the pull-backs from the common decision-making process can be seen all over Europe. In the UK political powers choose to leave the common decision-making process. Italy is openly defying the rules. Political powers in the government of Greece, Austria, the Czech republic, Poland and Hungary are all pushing against further integration and, in some cases, the full effect of EU law. The Courts in Denmark (openly; case), the Czech Republic (openly; case), and the Federal Republic of Germany (stealthily; case) already felt obliged to provide constitutional relief for extreme cases by limiting the full effect of EU law on their territory. And in the majority of EU Member States the Constitutional Courts signaled that if necessary they would do same in similar extreme cases where their constitutional identity or the core of the fundamental rights and liberties would be under threat.
The commitment device of treaty amendments is not only a way of raising the amount of common decision-making people sign up to, but also a way to lower the expected costs to an acceptable level after a shock in beliefs over these costs took place. The expected costs are indeed only an estimate and thus not fixed. They are based on the understanding of past experiences. If new experiences change the expected value of cooperation (for example after a visit of the Troika in Greece or an extension of the maturity of the loans granted by Germany), the willingness to undergo a common decision-making system will change. These expected costs can be lowered again by adapting primary and secondary law. This is however only the case if the Member States believe that the treaty will be binding.
If the treaty provisions are expected not to be credible, the adaptation of the rules is no longer a remedy to push back the expected cost of common decision-making when they rise. The amount of common decision-making or even the willingness to be a part of the common decision-making procedure will be lowered. Judicial lawmaking activities that lower the expected mitigating effect of treaty provisions, deprive politicians of an instrument that allows them to push the cost of common decision-making low enough for the peoples of Europe to be willing to participate in and strengthen the European integration.
The European Union
Member States that decide to leave the Union (like the UK) or not to join it (like Switzerland and Norway) will make the assessment that being part of the common decision-making process has a lower expected value than not joining. They choose not to participate in the Union at all. In the case of the EEA countries (like Norway), the cost seem to be lower if they limit the number of competences and the influence of the ECJ (through the EFTA Court). In the case of Switzerland even more limited cooperation is preferred.
If the ECJ is not able to effectively uphold the terms on which Member States agreed, they deprive politicians of an instrument to lower the costs of European integration, which can, in turn lower the amount of European integration for which people are willing to sign. This in the end lowers the chance of having an ever closer Union. Driving up the cost of common decision-making, it might instead achieve the dissolution of the Union.
The ECJ transformed the substantial constraints on the use of power by the ECB into formal constraints. It did this by adopting standards that the Court is unable to review, because those self-chosen standards depend on the assessment of facts they are unwilling to assess. In so far as those substantial constraints are a requirement for the political or legal constitutions to apply European law on their territory, the ECJ itself endangers the full effectiveness and uniform application of EU Law. In the long term it also deprives politicians of a trust-building instrument to overcome the cost of a common decision-making process, namely substantial treaty provisions of which the Member States can believe that they will be effectively enforced. In this way it raises the cost of common decision-making and will lower the expected value of participating in such a common decision-making process. This will slow down or even reverse European integration. All this runs against the basic ideas behind the EU, namely uniform application, full effect and ever closer Union.
 The blog post might give the impression that the goal of a provision in the treaty should not be considered at all by the ECJ. That impression would be wrong. The objection is not against the consideration of the goal that those who consented to the treaty amendment had in mind. That goal can be considered to give a sensible interpretation, to prevent circumvention and to prevent overgeneralization of the (always abstract) words used in a provision. The objection is limited to the practice that replaces a choice in a treaty by a more abstract goal behind that specific choice.
 It is true that the practice before the transfer of powers was differentiated over Member States. Although the criterion would have sufficed for all cases about the EMU that have arisen until now, it will indeed not help to distinguish what is and is not permissible where the ESCB is following a practice that was used in one Member State and not in another. That the practice is not a sufficient criterion to review all cases, does not mean that it should not be used in the first line of reasoning. The preparations of the statutes can be used in a secondary line of reasoning to find out what types of practice that differed between the Member States were intended to be allowed and prohibited.