The Constitutional Court's decision on pension reform: a good lesson in parliamentary law

Noëlle Lenoir, honorary member of the Constitutional Court, lawyer and member of the Scientific Committee of the Laboratoire de la République, looks back at decision no. 2023-849 DC of April 14, 2014 on pension reform. Calling the decision "a good lesson in parliamentary law", she discusses the constitutional issues that were debated.

Criticism of the decision has evolved over time

The Constitutional Court's decision on the amended Social Security Financing Act for 2023 reforming the pension system only disappointed those who would have liked to see it censured in its entirety. Having been unable to see the reform blocked by the procedural activism of certain members of parliament, they would have liked the Constitutional Court to cancel it outright. Did they really believe this? This is doubtful, given the criticism levelled at the Conseil before and after its decision.

These days, when a decision is disliked, it's the person making it who is personally targeted. The Constitutional Court was no exception to these ad personam attacks, which began several weeks before April 14th. Perhaps suspecting that the referrals made by senators and deputies (RN and NUPES in the National Assembly and left-wing groups in the Senate) had little chance of succeeding, it was necessary to discredit the decision-maker through its members: accomplices of those in power, insufficiently competent because they were not all qualified professors of constitutional law, or politicized because they were former ministers or prime ministers, from both the left and the right, it was necessary to cast doubt on the judge's impartiality.

Once the decision was adopted, possibly unanimously, the tone of the criticism changed. In contrast to the accusations of politicization, it was the "legalism" of the Constitutional Court that was stigmatized. According to its detractors, it should have reviewed the legislator's copy by "taking an interest in the economic and social aspects of the reform". In the absence of a qualified majority to adopt a motion of censure, theConstitutional Court was asked to censure the law.

And some called for "the transformation of the Court into a Constitutional Court", by which they meant that it could decide on its own merits. But what would have been said if the Constitutional Court had behaved like the US Supreme Court, whose politicization, especially since the legacy of the Trump era, is hardly a model?

On parliamentary procedure, a decision based on established case law

The referrals from the deputies and senators focused on the alleged misuse of procedure by the government and Senate authorities, rather than on the merits.

The question posed was whether the government, but also the Conférence des Présidents[1] and the parliamentary committees responsible for examining the text, had abused their rights by using the weapons at their disposal to counter parliamentary obstructionism.

To inflate the number of amendments and sub-amendments (over 30,000 in all), all kinds of methods were used: the same amendments were tabled on an individual basis, contrary to the usual practice of amendments by political group; hundreds of amendments provided for derogations profession by profession and distinct dates of entry into force measure by measure; speeches and points of order were counted by the hundreds, not to mention the systematic tabling of motions to refer each article of the text back to committee and the no less systematic requests for a vote by public ballot, and so on.

For the petitioners, the only way to comply with the Constitution was to let the debate drag on! The Constitutional Council's responses to the arguments put forward in the referrals are a real lesson in parliamentary law that all jurists and legislators should learn from.

On the choice of the rectifying Social Security Financing Act as the vehicle for reform.

Admittedly, the previous pension reforms in 2003, 2010 and 2014 were adopted within the framework of an ordinary law. However, in the eyes of the petitioners, the latter had the advantage of not being constrained, like the social security financing laws, by the pre-set deadlines of Article 47-1 of the Constitution.

However, the Constitutional Court pointed out that recourse to social security amending bills is not conditional, as they claimed, on "urgency, exceptional circumstances or a major imbalance in the social accounts".

The government could therefore choose the framework of the Social Security Finance Amendment Act for its reform, provided that, in accordance with articles L.O. 111-3-9 et seq. of the Social Security Code, it included "provisions relating to the current year that have an effect on the revenues of the compulsory basic schemes or the bodies contributing to their financing, and those relating to the base, rate and methods of collection of the contributions allocated to the compulsory basic schemes or the bodies contributing to their financing".

Concerning the application of the deadlines set out in Article 47-1 of the Constitution to the adoption of the Social Security Finance Amendment Acts.

According to this article, "if the National Assembly has not reached a decision on the first reading within twenty days of a bill being tabled, the Government shall refer the matter to the Senate, which must reach a decision within fifteen days...". This is what the government did; faced with a deluge of amendments and sub-amendments in the National Assembly, it sent the text to the Senate after twenty days.

As article L.O. 111-3 of the Social Security Code states that "the following shall have the character of a social security financing law: ... 2 ° Amending Social Security Financing Act", this procedure is applicable to both types of law, as the Conseil indicated.

On the use of Article 49-3 of the Constitution, which allows the government to commit its responsibility for a text before the National Assembly.

The petitioners argued that the government's commitment of responsibility under Article 49-3 could not relate to the text as a whole, and that it would have been necessary to hold a successive vote on the revenue and expenditure forecasts. A 1979 decision[2] had in fact censured the coming year's Finance Act on the basis of Article 40 of the Constitution, which requires a decision to be taken on the revenue side before voting on the expenditure side?

However, this rule, transposed to social security financing laws, does not override Article 45 of the Constitution, which allows the government, once a text has been drafted by the joint committee, to submit it to the assembly, free to accept or reject other amendments. The government could therefore engage its responsibility for the text as a whole.

The parliamentary petitioners could have anticipated this response, since it had already been given, notably in 2015[3], but above all in 2022 in connection with the Social Security Financing Act for 2023[4].

On the inadmissibility of amendments not previously submitted to the committee responsible for examining the text

Necessary for the "proper conduct of democratic debate", the right of amendment conferred on members of parliament and the government by article 44 of the Constitution is not exclusive of rules governing the organization of parliamentary debate. Thus, by virtue of the 2nd paragraph of article 44, "after the opening of the debate, the government may oppose the examination of any amendment which has not previously been submitted to the committee".

According to the Senate website, "this procedure is rarely used in practice, since the hypothesis of an amendment not being submitted to the committee is relatively rare". To prevent the law from being passed, parliamentarians have departed from this practice. They tabled thousands of sub-amendments "several days after the amendments" examined in committee were tabled.

The Constitutional Court simply ruled that, by declaring them inadmissible, the government had applied the 2nd paragraph of Article 44 of the Constitution, the purpose of which, it should be remembered, is to enable members of parliament to take a fully-informed position on proposed modifications to the text under discussion.

Recourse to a blocked vote

The petitioners argued that recourse to the blocked vote had altered the clarity and sincerity of the debate, a constitutional requirement whose principle was established by the Constitutional Court in 2005[5]. The aim is to ensure that the law is truly "the expression of the general will", in the words of article 6 of the 1789 Declaration of the Rights of Man and of the Citizen.

But once the debate has taken place and all opinions have been expressed, it must be possible to decide. This is the idea behind the 3rd paragraph of article 44 of the Constitution, which stipulates that "If the Government so requests, the assembly seized of the matter shall take a single vote on all or part of the text under discussion, adopting only the amendments proposed or accepted by the Government".

Unquestionably, in view of the numerous discussions and speeches during the session, the Constitutional Court considered, in accordance with its usual jurisprudence, that "the blocked vote procedure did not have the effect of preventing discussion of the provisions on which the Senate was asked to take a single vote".

On the implementation of Senate rules concerning the closure of debates, the determination of an order of priority for amendments and the examination of their admissibility.

This time, the criticism was levelled not at the government, but at the Senate itself. As a result, the Constitutional Court - in a rare move - asked the Presidents of the two assemblies for details of the procedure. The President of the Senate thus indicated that "the rate of inadmissibility for the text as a whole was around 48%, which is ultimately not much higher than the rate recorded during the examination of the social security financing bills for 2022 (41%) and 2023 (38%), despite the massive submission of sub-amendments which were the subject of specific declarations of inadmissibility".

Three criticisms were levelled at the way in which the debate on Article 7 of the bill, which raised the retirement age from 62 to 64, was conducted:

   - the Senate should not have resorted to its rules of procedure to limit the general discussion on the article to two speakers of opposing views; but as the Constitutional Council noted, there had already been 64 speeches on this article!

   - the definition of an order of priority for the consideration of amendments led to the collapse of 1,300 amendments; but these were incompatible with amendments already adopted!

    - thousands of sub-amendments were declared inadmissible, but they all contradicted the amendment to which they were supposed to relate!

On the cumulative use of procedural means to avoid blocking debate.

You didn't need to be a rocket scientist to know that this argument didn't hold water, especially as the issue was not really new.

In its 2006 decision on the "late" contrat première embauche (CPE), which had also aroused strong protests on certain benches of the Chamber, the Constitutional Court had considered that "the fact that several procedures provided for by the Constitution were used cumulatively to speed up the examination of the law referred to is not in itself such as to render unconstitutional the entire legislative procedure leading to its adoption". He also ruled that "the combined use of the various provisions set out in the Senate's rules for organizing the exercise of the right of amendment can no more have the effect of rendering the legislative procedure contrary to the Constitution"[6].

So there you have it.

As for the other arguments in support of the insincerity of the parliamentary debate, the Constitutional Court found that the documents attached to the bill complied with the provisions of the Social Security Code, and rejected the unsubstantiated argument that the government had acted in bad faith and "distorted the broad outlines of the balance of the social security system".

An equally predictable decision on the merits

Curiously, the eventual disappearance of the special schemes7 did not attract any criticism, as if the principle were inescapable. The constitutionality of only two articles was invoked.

The first concerned article 10, on the postponement of the retirement age and the acceleration of the increase in the length of contributions.

According to the petitioners, the postponement of the retirement age and the acceleration, decided in 2014, of the length of contributions required to qualify for a full-rate retirement pension to 43 years, would call into question our social protection system and therefore violate paragraph 11 of the 1946 Constitution, which states that the Nation "guarantees to all, in particular (...) to workers..." this protection.

For the NUPES deputies, the legislature must be forbidden to roll back any social advances.

In a 1984 decision[8], the Court stated that, with regard to freedom of the press, "as this is a fundamental right, the law can only regulate its conditions with a view to making it more effective or reconciling it with other rules or principles of constitutional value". But this formula, never applied to social rights, did not stand up to scrutiny. Since then, the Constitutional Court has reiterated that the legislator is free to amend the law, provided that it does not "deprive constitutional requirements of legal guarantees". The idea behind this cryptic formula is that no one has the right to maintain legislation unless the new law does not disregard constitutional principles.

Thus, there is no constitutional requirement for the absolute preservation of "acquired rights". When legislation evolves, it is often to reconcile contradictory values. Such is the case with pay-as-you-go pension schemes, where the need to take account of demographic trends and the need to ensure the financial equilibrium of the social security system must be taken into account. Yet, as the April 14 decision states, the legislator's objective "to ensure the financial equilibrium of the pay-as-you-go pension system and thus guarantee its sustainability" does not call into question the "national solidarity in favor of retired workers" that it aims to preserve over time.

Some law professors have deplored the fact that the Constitutional Court did not question the relevance of this objective. But what would they have said if, during the examination of the "35-hour" law in 1998, as requested by the parliamentary petitioners at the time, the Court had annulled the law by denying any validity to its "objective of reducing unemployment and safeguarding employment" then invoked by the government!

The Constitutional Court does not judge on the basis of opportunity, and it has therefore verified, in accordance with its constant approach, that the measures taken in the law are not "inappropriate" with regard to the objective of financial equilibrium of the pension scheme that the legislator has set itself.

Concerning article 11, which sets the early retirement age for employees who started work before the age of 21.

The criticism levelled at this article may seem surprising, since the provision in question was precisely intended to allow employees with a "long career" to take early retirement.

The argument that this violated the principle of equality was easily dismissed as an error of interpretation, as the provision had "neither the purpose nor the effect of extending the insurance period for people who began working before the age of 21 beyond the total insurance period required of other insured persons".

Similarly, the Court noted that the fact that some employees with long careers had to contribute after reaching the age set for their early retirement was in line with the logic of the overall system of retirement by distribution.

Finally, the decision of 14 April annulled, as expected, several "social riders", i.e. provisions that cannot be included in a social security financing law because they do not affect its financial balance.

A classic decision, to conclude, which should at least put an end to the controversy over the integrity of parliamentary procedure. Apparently, the only remaining debate concerns the use of Article 49-3 on the government's commitment of responsibility for a text.

If the time has passed when Michel Rocard, Prime Minister, used it 28 times during his three-year term of office, it is advisable to think carefully before deciding to do without this "anti-filibustering" weapon.

Let's not forget: this procedure was wanted by all party leaders, left and right, when the 1958 Constitution was drafted. And for good reason: the multiplication of "questions of confidence" in the Senate as well as in the Chamber of Deputies had generated a chronic ministerial instability which Article 49-3 was able to remedy.

There are two main reasons for wondering whether this procedure is not still specially adapted to the French context: firstly, our allergy to any kind of reform, in the name of safeguarding what we call "acquired rights", which is in fact often detrimental to solidarity; secondly, the fragility of political alliances, which means that parties in France, unlike in other major democracies, are ephemeral. This is evidenced by the virtual disappearance of the parties that held sway under the Third and Fourth Republics and the weakness of those that founded the Fifth Republic.

It's not enough to jump up and down shouting "Sixth Republic, Sixth Republic, Sixth Republic" to know where we're heading. The most important thing is not 49-3, but safeguarding our democracy. The 1958 Constitution has been revised twenty-four times. A pause would be salutary.

[1] The Conference of Presidents in the National Assembly and in the Senate is composed of the bureau of the assembly, the presidents of the committees and the presidents of the groups, as well as the general rapporteurs of the finance committee and the social affairs committee. The government may delegate a representative, usually the Minister responsible for relations with Parliament. The role of this body is to set the program of parliamentary business, which is published in the Journal Officiel. In the Senate, it can determine the overall amount of time available for debate by speakers from the various groups. In the National Assembly, it can set the maximum duration for the examination of a text as a whole.

[2] Decision no. 79-110 DC of December 24, 1979.

[3] Décision n°2015-715 DC du 5 août 2015, Loi pour la croissance, l'activité et l'égalité des chances économiques.

[4] Decision no. 2022-845 DC of December 20, 2022.

[5] Decision no. 2005-512 DC of April 21, 2005.

[6] Decision no. 2006-535 DC of March 30, 2006 on the Equal Opportunity Act.

[7] It concerns employees of the electricity and gas industries, the RATP and the Bank of France, as well as clerks and employees of notaries and members of the Economic, Social and Environmental Council.

[8] Decision no. 84-181 DC of October 11, 1984, Loi visant à limiter la concentration et à assurer la transparence financière et le pluralisme des entreprises de presse.