The Void Ab Initio Theory in Comparative Perspective: J Marshall, H Kelsen, and Beyond

1 Introduction

The void ab initio theory is traditionally associated with the retroactive effect of an unconstitutional statute in diffused systems of judicial review; as opposed to the prospective effect of unconstitutionality in Austria, the archetype model of centralized judicial review. Two of the most important recent judgments in Europe illustrate the significance of the matter. The UK Supreme Court ruled in R (Miller) v The Prime Minister Cherry v Advocate General for Scotland ^[1] that the prorogation of Parliament, in the course of Brexit, was founded on ‘unlawful advice,’ which was ‘outside the powers of the Prime Minister’:

69. […] This led to the actual prorogation, which was as if the Commissioners had walked into Parliament with a blank piece of paper. It too was unlawful, null and of no effect. 70. It follows that Parliament has not been prorogued and that this court should make declarations to that effect. […] [Ι]t appears to us that, as Parliament is not prorogued, it is for Parliament to decide what to do next.

The ‘prorogation’ was void ab initio, and, for legal purposes, it is as if it had never existed. By contrast, notwithstanding the general rule of void ab initio in German law,^[2] the Federal Constitutional Court of Germany ruled in the European Central Bank ultra vires case on the PSPP, that ‘[the] judgment [of the CJEU] constitutes an ultra vires act that is not binding upon the Federal Constitutional Court’ [para 154], which ‘must conduct its own review to decide’ whether the Eurosystem’s decisions on the adoption and implementation of the PSPP remain within the competences conferred upon it under EU primary law [para 164]. As these decisions lack sufficient proportionality considerations, they amount to an exceeding of the ECB’s competences [paras 232, 234].

Accordingly, the Court set the time of the effects of its decisions for the future as following:

6. […] Following a transitional period of no more than three months allowing for the necessary coordination with the ESCB, the Bundesbank may thus no longer participate in the implementation and execution of the Decisions [at issue], […] unless the ECB Governing Council adopts a new decision that demonstrates in a comprehensible and substantiated manner that the monetary policy objectives pursued by the ECB are not disproportionate to the economic and fiscal policy effects resulting from the programme. […] [para 235]^[3]

The contrast between the void ab initio doctrine and the effects of unconstitutionality for the future is not a narrow technical matter concerning the temporal aspect of judicial review. Rather, it is associated with deeper issues regarding the role and the nature of judicial review of the constitutionality of legislation. Section 2 of this paper discusses these differences as explicated in the decision of Chief Justice J Marshall, in Marbury v Madison,^[4] and in the writings of H Kelsen, the father of the Austrian (centralized) system of judicial review. The American system embraced the void ab initio doctrine, or the ex tunc effect of unconstitutionality, while the Austrian one, the pro futuro, or ex nunc effect of unconstitutionality. Despite this theoretically imbued discussion, Cappelletti and Cohen^{[5] , [6]} note that both systems tend to converge on the basis of practical considerations.

Section 3 enquires into the factors that explain why the temporal effect of judicial decisions, retroactive or retrospective, is left to the control of the courts. These considerations are associated with the question whether the void ab initio doctrine is constitutionally mandated, or not. The answer is negative; however, although prospectivity is allowed, the retrospective effect of void ab initio is founded on strong constitutional principles, associated with the protection of rights and effective judicial relief; and it can also be mandatory, eg in the context of substantive criminal law.

In order to organize the discussion, this paper introduces the distinction of the effect of unconstitutionality in terms of:

1. the type of the effect, as being primarily one of ‘legal sanction’ or rather one of ‘policy dictates;’

2. the grounds for the effect, depending on the degree of ‘legality’ of the court’s ruling, on the basis of prior law;^[7] and

3. the scope of the effect, depending on the ‘adjudicative’ or ‘polycentric’^[8] nature of the problems tackled. So, though both the English and the German court find violations of the rules of competence, the former sanctions a specific past incident, while the latter refers to future action concerning participation in the implementation and execution of ECB decisions, ie, complex issues of public policy. The former is devoted to a discussion of core problems of constitutional law, eg, prerogatives, sovereignty of parliament, prorogation, accountability. The latter is immersed in highly technical issues over the distinction and preponderance of economic, fiscal and monetary aspects, reviewing ECB’s actions. Finally, the issue in the English case was clearly justiciable, while in the German one it was polycentric par excellence, implicating ‘complex repercussions’^[9] of ECB’s decisions. The void ab initio theory is most adequate for legal sanctions, on legal grounds, over adjudicative facts. ^[10]

Section 4 discusses the practically very important consequence of the void ab initio theory of revival or not of preexisting law, as if the latter had never lost its validity, once the unconstitutional statute is declared null and void. This matter is particularly important depending whether the theory is applied to individual, negative, rights or to social, positive, rights. Application of the void ab initio theory in the context of positive rights encroaches on legislative competence and violates the separation of powers (Section 5).

The discussion of the void ab initio rule shows its concern for the fidelity to law, the supremacy of the Constitution and the rule of law.⁶ It reveals, however, the risks for abuse of the rule in violation of the separation of powers, particularly in the field of social rights, in favor of judicial supremacy.

2 Marshall and Kelsen on the Temporal Effects of Unconstitutionality

The void ab initio theory goes back to Marbury v Madison (1803). Chief Justice Marshall wrote^[11] that:

[A] legislative act contrary to the Constitution is not law; […] an act of the Legislature repugnant to the Constitution is void. […] If two laws conflict with each other, the Courts must decide on the operation of each. […] Those, then, who controvert the principle that the Constitution is to be considered in court as a paramount law are reduced to the necessity of maintaining that courts must close their eyes on the Constitution, and see only the law. This doctrine would subvert the very foundation of all written Constitutions. It would declare that an act which, according to the principles and theory of our government, is entirely void, is yet, in practice, completely obligatory.

Another well-known formulation of the void ab initio rule is found in Norton v Shelby County: ^[12] ‘an unconstitutional act is not a law; it confers no right; it imposes no duties; it affords no protection; it creates no office; it is, in legal contemplation, as inoperative as though it had never been passed.’ As Oliver Field wrote, in 1935, the ‘theory that an unconstitutional statute is void ab initio is the traditional doctrine of American courts as to the effect of an unconstitutional statute.’^[13]

The void ab initio theory is the result of the logic of validity. An invalid law is null and this means that as a matter of logic it is as if it never existed. As Kelsen clarifies, ‘An act that is null lacks any legal character from the beginning, so that there is no need for another legal act to strip it of its pretended legal quality. If such a second legal act is required, we are faced with annullability, and not with nullity.’^[14]

In its turn, the logic of validity presupposes the existence of superior law over subordinate law, so that the former takes precedence and sets aside the latter, if they are in conflict. This is the relationship of the Constitution to the statute. From this flows the nature of the judicial duty: not to ‘pronounce a new law, but to maintain and expound the old one,’ according to Blackstone, who stated the rule of common law. ‘The judge, rather than being the creator of the law, was but its discoverer.’^[15] The Constitution is preexisting fundamental law and binds the judge who upholds it over the statute.

Marshall airs the same view, echoing the Madisonian model of checks and balances:

It is emphatically the province and duty of the Judicial Department to say what the law is. […] If two laws conflict with each other, the Courts must decide on the operation of each. […] This is of the very essence of judicial duty.^[16]

When Marbury v Madison was decided, the notion of firm and solid constitutional law, ie, the idea of the constitution as ‘hard law,’^[17] was embedded. As Field noted, ‘the emphasis laid at that time upon guarantees against arbitrary action on the part of government contributed much to this idea, both as those guarantees related to property rights and as they related to personal rights closely associated with the idea of freedom: political and legal freedom.’^[18] The common law offered the background for the constitutional interpretation of concepts like liberty and contract.^[19] The contract clause was violated when legislation interfered with contractual relations prior to its enactment;^[20] such legislation was considered to be retroactive, and the void ab initio rule was the adequate remedy to undo this retroactivity, by holding the law null and invalid, leaving thus private contracts unimpaired.

The most demanding conditions of legality being satisfied, the legal sanction in the form of judicial review of legislation was beyond question, because the material unconstitutionality was naturally merged into the formal one. Far from being merely a linguistic trick, the understanding of material unconstitutionality in terms of violation of the limited powers of government, in terms of exceeding legislative competence,^[21] was inherently connected with the need of judicial review to stand on safe ground in order to consolidate its legitimacy.^[22] In the American system of separation of powers, none of the three departments should ‘overstep those limits that had been placed upon all government to protect individual rights;’ ‘excesses of power should not take place under this system,’ and so the void ab initio doctrine assured that ‘legislative excesses were to leave no effect on the law.’^[23] As already said, the void ab initio rule is most adequate for legal sanctions, on legal grounds, over adjudicative facts.

Expressing material unconstitutionality in terms of violation of formal rules of competence is a move of judicial prudence, a sign of self-restraint, and a declaration of respect for legality. It is not accidental that in the historical evolution of the petition for the annulment of administrative acts, in France, the lack of competence developed first.^[24] A most severe case of lack of competence is ultra vires; and it is, therefore, not hard to understand why the German Constitutional Court invokes ultra vires in order to question the legality of ECB’s actions. But ultra vires is only half of the equation; what is missing from the German decision is the other critical part, ie, firm legal judgment on adjudicative facts.^[25]

The void ab initio theory has its own presuppositions and assumptions. It is not merely a theory on the temporal effect of the judgment of unconstitutionality. It is part of a broader judicial ethos.^[26] Moving away from the void ab initio rule marks a similar change of the prevailing judicial ethos. The following is one of my favorite phrases, because it captures powerfully this gradual shift:

The process by which the limits of constitutional authority have come to be made more and more to depend on proof of facts has gone forward so gradually that only recently has it attracted attention. It is probably not too much to say that such a yard-stick of constitutional power would have been abhorrent to Marshall and his contemporaries. For them authority in a governmental organ existed or did not exist.^[27]

But, of course, the void ab initio theory is not the only game in town. In the USA, Linkletter v Walker (1965) was the first case to limit the effect of unconstitutionality, prospectively. The Supreme Court stated in Linkletter ^[28] that:

Austin maintained that judges do, in fact, do something more than discover law; they make it interstitially by filling in with judicial interpretation the vague, indefinite, or generic statutory or common law terms that alone are but the empty crevices of the law. Implicit in such an approach is the admission when a case is overruled that the earlier decision was wrongly decided. However, rather than being erased by the later overruling decision, it is considered as an existing juridical fact until overruled, and intermediate cases finally decided under it are not to be disturbed.

Like Austin, Kelsen subscribed to legal positivism and espoused similar doctrines. Kelsen rejected the void ab initio interpretation regarding the effect of the Supreme Court’s decision on the unconstitutional statute. That would mean that the decision ‘annuls the statute generally and with retroactive force so that all the legal effects which the statute had before are abolished.’^[29] As a consequence, in his view, the void ab initio rule violates the prohibition of ex post facto laws of the American Constitution.^[30] Kelsen argued that a legal act ‘is not void, it is only voidable’. ‘Within a system of positive law there is no absolute nullity. It is not possible to characterize an act which presents itself as a legal act as null a priori (void ab initio).’^[31] The reason is because ‘before this declaration the act is not null, for being “null” means legally non-existent. And the act must legally exist, if it can be the object of a judgment by an authority.’ ‘The act is “null” only if the competent authority declares it null’.^[32] This is reflected in the Austrian Constitution of 1920 (as amended, art 140 para 7).

Kelsen insisted on the creative aspect of the judgment of unconstitutionality, which produced legal effects: ‘From a legal point of view only the opinion of the court is decisive. Therefore the statute must be considered valid so long as it is not declared unconstitutional by the competent court. Such a declaration has, therefore, always a constitutive and not a declaratory character.’^[33] Kelsen accepted the legislative nature of the decision of unconstitutionality, being actus contrarius. ^[34]

By contrast, the void ab initio theory can be analyzed into a legal fiction (the statute is void, as if not passed), combined with a temporal legal effect (ab initio). The legal fiction is a counterfactual, which is nevertheless assumed and accepted as true, because it serves important principles or purposes of the law. As Kelsen noted, ‘The juridic fiction can only involve a fictitious legal claim, and not a fictitious actual claim.’^[35] ‘The application of law, just as the creation of law, does not really intend the cognition of law, but its realisation, it is about acts of the will.’^[36] The purpose of the legal fiction served by the void ab initio theory is the supremacy of the Constitution. Kelsen’s argument that the void ab initio temporal effect runs against the prohibition of ex post facto legislation can be refuted. Allegedly, the void ab initio effect does not present a case of strong retroactivity, but of a weak one, to the extent that it operates retrospectively,^[37] having the character of judicial relief – of a legal remedy^[38] – being beneficial^[39] for the citizen. This is consistent with the principle emphasized in Marbury v Madison that there can be no legal right without a remedy.^[40]

Relevant to the previous discussion is the relationship between validity and authority. Defending the declaratory character of the judgment of unconstitutionality tends to efface authority and have it absorbed by validity. Accepting the constitutive character of the judgment in question reveals the reverse tendency. But as already said, it is not on the ground of logic that this issue is to be decided. The issue is rather one of purposive interpretation of substantive law. The relationship between validity and authority in the exercise of judicial review depends on the existence of legal sanctions, on legal grounds, over adjudicative facts. The question of retroactivity is intertwined with that of the nature of judicial decision-making, and the opposition between old law (retroactive) v new law (prospective). However, one must not exaggerate the practical difference between Marshall and Kelsen. After all, Marshall is the author of the emblematic invitation and reminder, that ‘it is a Constitution we are expounding.’^[41] In an age when a great part of constitutional law is case law, the interpretive role of the judge is well embedded, and inescapably many cases are the outcome of authority, rather than of preexisting validity. This is certainly a matter of legitimacy, but not a straightforward one. It is not always easy to say when a new rule falls into the one category or the other; a new rule does not necessarily mean new law; it may reinterpret an old, but imperviously applied legal principle, correcting that misapplication. Brown v Board of Education,^[42] which overruled Plessy v Ferguson,^[43] declaring the unconstitutionality of the ‘separate but equal’ doctrine is emblematic. Brown is certainly new law, seen flatly; but, in the light of Justice Harlan’s famous dissent in Plessy v Ferguson (‘Our constitution is color-blind, and neither knows nor tolerates classes among citizens’),^[44] if one thinks in terms of validity, pondering the gap between the normative and the descriptive, the first impression may look deceptive, because the old law was simply wrong law, while the new law is old principle correctly applied.

3 Between Retrospective and Prospective Effect of Unconstitutionality

O Field has underlined since 1935 the impact of pragmatic considerations in avoiding inflexible dogmatic solutions to the problem of temporal effect (ex tunc or ex nunc) of the judicial decision declaring a statute unconstitutional. The void ab initio rule is not universal, because ‘there are several rules or views, not just one, as to the effect of an unconstitutional statute. All courts have applied them all at various times and in differing situations.’^[45] Field discussed the presumption of constitutionality of the law, according to which ‘a statute is valid until it is declared by a court to be invalid,’^[46] and case-to-case judicial review, which holds that the unconstitutional statute ‘is neither wholly valid nor wholly invalid,’^[47] depending on the circumstances, and as a result ‘no sweeping law-making effect is attributed by this view to judicial decisions.’^[48] The advantage of this approach is that ‘it permits of great elasticity in the law,’ but the great disadvantage is that ‘it does not lend itself to the formulation of rules of law. Prediction becomes well-nigh impossible.’^[49]

The judicial discretion to apply one or another of the aforementioned views is founded on the assumption that the void ab initio rule is not constitutionally required. Indeed, in Great Northern Railway Co v Sunburst Oil & Refining Co,^[50] J Cardozo, who favored the pro futuro effect, accepted for the Court that the American Constitution has no voice upon the subject and as a result, a state can define the limits of adherence to precedent either way (ex tunc, or ex nunc).

Turning to the Austrian centralized system of judicial review, in X v Austria,^[51] in a case alleging the violation of art 1 of the First Protocol (private property) because the Constitutional Court of Austria restricted the effects of the annulment ‘to matters subsequent to the entry into force of the annulment,’ the European Commission of Human Rights held that: ‘There is no provision in the Convention forbidding the High Contracting Parties to annul a statute ex nunc and nevertheless to continue to apply it to matters prior to the annulment.’^[52]

In American constitutional law, Thayer’s classic argument^[53] in favor of the presumption of constitutionality of statutes has several common elements with Kelsen’s thought. Thayer argued that the mistake of the legislature must be ‘a very clear one,-so clear that it is not open to rational question,’^[54] because ‘much which will seem unconstitutional to one man, or body of men, may reasonably not seem so to another.’^[55] Here is found the element of discretion looming in the interpretation and application of norms, and as a result the constitutive character of the judicial decision declaring the unconstitutionality of the law. ‘The judges were allowed, indirectly and in a degree, the power to revise the action of other departments and to pronounce it null.’^[56] Implicit here is a conception of unconstitutionality whose focus is on the notion of competence: ‘The judicial function is merely that of fixing the outside border of reasonable legislative action, the boundary beyond which […] legislative power in general, cannot go without violating the prohibitions of the constitution or crossing the line of its grants.’^[57]

An important problem concerns the prospectivity of the judgment of the highest courts in a system of case-to-case judicial review, where there is no doctrine of binding precedent, as in the USA, and jurisprudence is not recognized as a formal source of law. This is the case of Greece, a country in the Civil Law tradition. One could argue^[58] that in a true system of decentralized review incidenter, ‘it is inconceivable’ to talk about limiting the effects of unconstitutionality, because the consequences of the judgment are exhausted, by definition, to the litigants of the specific case, and to no other person, or court of law. The effect of the decision binds the parties (res iudicata facit ius inter partes) and in the specific case it is necessarily retroactive; this has been also accepted in the Austrian model since 1929, as an incentive and reward to bring the case to the court. Further debate over the retrospective or prospective effect is redundant.

However, the decision of unconstitutionality of the Greek Council of State (the supreme administrative court) clearly exerts an interpretive influence and plays a guiding role^[59] which differs from res iudicata because it concerns different future legislation with similar, however, defects to the previous one. Here the crucial questions are: the similarity of the future case to that already decided; and the level of abstraction and inclusiveness in the previous decision of the Council of State, regarding the interpretation and the application of the Constitution to the litigated facts.^[60]

In a recent judgment the Auditor’s Court^[61] (one of the three supreme courts of Greece) ruled that the Plenary Session of the Court, when deciding (according to art 100 para 5 of the Greek Constitution) over a preliminary question of constitutionality, is ‘free from procedural constraints to hear all crucial argumentation, as well as legislation and case law, even if subsequent’ to the applicable law in the litigated case. Otherwise, the Plenary Session would run the risk to render an ‘unpersuasive decision,’ instead of one ‘binding, not for its formal authority, but on substantial grounds.’ Therefore, the decision-time of the court is controlling, not the time of the litigated acts, or of the transactions.

Regarding the Austrian model, Kelsen argued that: ‘Out of consideration for the ideal of legal security, one should, in general, make the cassation of a general legal norm on the ground of legal defectiveness effective only pro futuro.’^[62] Judicial discretion in setting the time of the effect of unconstitutionality is today more or less generally recognized, either in principle as a matter of inherent judicial power, or for practical purposes in order to remove unpalatable consequences of the decision. This power is more or less equivalent to that exercised by the courts in the course of interim relief, ordering injunctions or mesures conservatoires.

If the practical limit to the implementation of the void ab initio theory is set by the legal maxims lex non cogit ad impossibilia, ad impossibilia nemo tenetur, then, balancing the interests at stake seems advisable in deciding the ex tunc or ex nunc factor. Relevant criteria^[63] for the exercise of such discretionary power may include the protection of legal rights and interests of the immediate parties to the dispute, prudential considerations facilitating the effectiveness of the judicial decision, the avoidance of a flow of litigation which could unduly burden the courts, or the prevention of significant costs to the public purse with negative macroeconomic consequences etc.

In an important article on the issue of retroactivity, Samuel Beswick^[64] analyzes five frameworks used by the US Supreme Court in the effort to grapple with the problem of retroactivity (the philosophical one regarding the nature of the judicial duty between lawmaking and law-applying; balancing of interests, in the light of the legitimate expectations of the parties; the choice of law as in conflict of laws; the availability of remedies depending on the predictability of the new law; or the application of prospectivity as an exception). The author develops a sixth framework, that of the ‘right of action,’ whose focus is on the justiciability of claims founded on the newly introduced law.

Beswick’s model has a strong procedural aspect and depends on specific arrangements regarding the discretion allowed by procedural law to the judge of constitutionality, among the possibilities of: a) retroactivity, b) prospectivity from the time of publication of the judgment of unconstitutionality, and c) postponement of the effects of the judgment further away from the decision. The procedural approach focusing on issues of justiciability and on conditions for legal actions is the most adequate to tackle the problem of retroactivity, confirming Cappelletti and Cohen^[65] who noted some time ago that both systems, the American and the Austrian, tend to converge on the basis of practical considerations. The void ab initio theory is at the intersection of substance and procedure; the void is a matter of substantive law, the ab initio depends on existing procedures.

Indeed, the notion of the effects of the decision is multifaceted. The effect of unconstitutionality must be synchronized temporally, harmonized systematically, and unified spatially.^[66] These dimensions are naturally under the immediate control of the court, being effects of the court’s decision; thus, the court itself should assess them. So, the French Constitution (art 62, s 2), regarding the ‘priority question of constitutionality’ submitted by the courts to the Constitutional Council in the course of litigation,^[67] allows for immediate or future effect, while the main trial is ordered to stay. Furthermore, there are also hard structural aspects diversifying the effects of the decision. Between the litigants the doctrine of res judicata hampers reopening the issues; and if the court adopts a strictly narrow holding, limited among the parties, then, the void ab initio rule is more or less inescapable; because it would seem contrary to justice and fairness, and also absurd, to insist on the ex nunc effect and not to recognize the retroactivity of the decision for the immediate parties to the dispute, who brought the case to court, asking for judicial protection. But it is difficult to limit the effects of the decision to the parties only. As already said, the range of interests in the processes of balancing may include the coherence of the legal order, guidance to the lower courts, legal certainty, prevention of a flood of trials etc. The complexity of the effects explains why the judicial control over the effects of the decision is an inherent power of the courts and is recognized to be so across the various systems of judicial review. This is particularly true of the highest courts, which usually exercise control over their own procedure.

4 The Consequence of the Void Ab Initio Theory: The ‘Awakening’ of Old Law

The main consequence of the ex nunc unconstitutionality of the law is in Kelsen’s system simple and straightforward. ‘The legal effect of the invalidation of conflicting norms [is] in accordance with the principle lex posterior derogat priori.’ ‘This means that it is not at all the case that the legal situation which existed before the annulled statute entered into force automatically comes to life again with the cassation of a statute on the part of the constitutional court.’^[68] As Kelsen believed,^[69] a constitutional court can be empowered to either fill the legal vacuum created by the annulment, or to defer to the legislature.^[70] This happens in France.^[71] According to Kelsen: ‘It is advisable to leave it to the discretion of the constitutional court to decide in which cases it wants to make use of this authorization to resurrect the old legal situation.’^[72] The wisdom of Kelsen’s advice over the risk of reviving old law wholly inadequate for changing circumstances is confirmed by recent Greek developments in the field of judicial review. The void ab initio theory requires by inexorable logic to ‘undo’ the effect of the unconstitutional statute and to turn back the clock. The old law does not become dead letter by an unconstitutional statute; rather, it is in a state of hibernation, waiting to awake as soon as the winter of unconstitutionality is set aside by the spring of judicial review.

The void ab initio theory is firmly established in Greek constitutional tradition. Its canonical expression is given by N N Saripolos in his System of Constitutional Law of Greece (1923). Saripolos, a major scholar learned in French and German law, with significant role in the osmosis of German legal science with the French tradition, wages a war on two fronts: on the one hand, he is adamant that an unconstitutional law ‘is not law, in any sense of the word,’^[73] and the courts have ‘the right and the duty to interpret and apply the Constitution, which is the fundamental law, not the unconstitutional law;’^[74] on the other hand, in contrast to current attitudes, which highlight the convergence of the models of judicial review,^[75] he argues vehemently that Greek incidenter review is ‘completely different’^[76] from Austria’s centralized system, because the Austrian Constitutional Court has the power to abolish the law by undertaking the authentic interpretation of the Constitution;^[77] on that occasion, it exercises in essence constituent power.^[78] Saripolos defends the supremacy of the Constitution, which depends on the distinction of constituent from constituted power, and, by the same token, the separation of powers, which is violated if the court imposes the principle of judicial supremacy.^[79] Saripolos, therefore, agrees with Kelsen’s defense of judicial review, but he disagrees with the judicial power to abolish the unconstitutional law, because Saripolos accepts, like Carl Schmitt,^[80] that the power to abolish the law involves the authentic interpretation of the Constitution by the judiciary.

The void ab initio theory was reaffirmed in Greece in the 1960s: Quod nullum est, nullum producit effectum.^[81] But since 2016, it has become the catalyst for the transformation of judicial review in the country to the point of changing league, from ‘weak type’ to ‘strong type,’ characterized by ‘normative finality,’^[82] to use Tushnet’s terminology. In 2012, in the course of the Greek sovereign debt crisis, as part of the obligations of the second Memorandum of Understanding between Greece and its creditors, further cuts of salaries for special categories of civil servants (eg army personnel, university professors, etc) were legislated, implementing severe austerity measures. In 2014, the CS found them unconstitutional. The executive was under strict obligation, imposed by GRCon art 95 para 5, to comply with the judgment of the CS annulling the administrative acts, which implemented the cuts. As a result of the void ab initio theory, the law prior to the cuts of 2012 (antecedent to the economic crisis), was resurrected.^[83] As compliance was pending, the legislature passed a new law, with retroactive effect, to return only half of the cuts of the previously given pensions, on the grounds of public interest and budgetary limitations.

In the decision 14/2013 the Special Supreme Court (which is a Court for the resolution of conflicts resulting out of opposite judgments over the constitutionality of a provision rendered by any two of the three supreme courts of Greece) ruled that the doctrine of res judicata and the principle of separation of powers did not impede the legislator from regulating with general and abstract laws issues which had been litigated and decided by the courts. Greek public law did not recognize a strong protection of vested rights. However, the CS, following the void ab initio logic, ruled that the legislature circumvented the constitutional obligation of the executive to comply with the Court’s judgments, because the judgment of unconstitutionality had the automatic effect of resurrecting the previous law, which had never been lawfully abrogated: the unconstitutional statute never existed, for legal purposes.^[84] The void ab initio theory turned the constitutional obligation of the executive to comply with the CS’s judgment of unconstitutionality into an obligation of the legislature! Partial compliance was equal to legislative abuse, dodging the constitutional obligations of the executive. As a result of the void ab initio theory, the problem of applicable law to replace the unconstitutional statute disappears ‘as if by magic;’ a mirage for those who do not understand the logic of legal validity. The fact that the legislature had twice made clear its intention to abolish the resurrected law, first by the statute found unconstitutional and next by the statute enacted following the unconstitutionality, which returned half of the cuts, proved legally irrelevant for the void ab initio theory. But dogmatic automations do not solve complex legal problems, generated by hard economic facts.^[85]

The CS backed off from the rigid view of the void ab initio effect, allowing for greater flexibility of the legislature in modifying the resurrected legislation, after the judgment of unconstitutionality.^[86] But the leniency of the approach is less important than the asserted constitutional basis: the Court reviewed such cases as falling under the judicially created novel obligation of the legislature to comply ^[87] (by not circumventing the main obligation of the executive), instead of applying the basic provision of judicial review incidenter. One can hardly exaggerate the distance between the two positions. The latter subjects the new law to judicial review, as any other piece of legislation, as long as the rule of res judicata is respected, without any violation of the impartiality and generality of the law. By contrast, the invocation of the constitutional obligation of the legislature to comply with the judgment of the CS on the unconstitutionality of the statute introduces a novel principle in Greek judicial review, similar to the French autorité de la chose interprétée. ^[88] This principle, when applied to a substantially different statute from the one previously held unconstitutional, amounts to a judicial claim addressed to the legislature to comply (art 95 para 5 GRCon) with the constitutional interpretation of the CS, instead of merely legislating in conformity with the Constitution (art 94 para 3 GRCon)! But the former is the gist of judicial supremacy in as much as the court becomes the Constitution’s authoritative^[89] interpreter, whose word binds the other branches of government. At the root of these developments is the void ab initio doctrine, which verifies Field’s statement that: ‘Most of the bad law in the decisions in this branch of the law has come from the inability of some courts to free themselves from the overpowering force of the logic evidenced in the void ab initio rule, once its premises are assumed.’^[90]

5 The Void Ab Initio Theory Between Positive and Negative Rights

The most problematic aspect of the void ab initio theory is not the ex tunc effect of the unconstitutional law, but the resurrection of the previous legislation to replace the invalid one. However, these two questions are different and distinct. The latter is one step further away from the temporal effect of the invalidity and concerns the applicable law, following the judgment of unconstitutionality. Why, then, do they tend to be assimilated? Because the void ab initio theory, in both prongs, is fine and fully consistent with the separation of powers and the judicial role in the context of individual rights and negative liberty. Operating as a shield against the state, individual rights render illegal a statute from the time of its enactment. Besides, no drastic intervention of the legislator is required. The court declares the law invalid and inapplicable to the litigated legal dispute and sets the accused free.^[91] The case of individual rights against the state is most likely to involve legal sanctions, on legal grounds, over adjudicative facts; and the issue of resurrecting old law is secondary.

Substantive criminal law is the privileged field of application of the void ab initio rule. ‘An unconstitutional law is void, and is as no law. An offence created by it is not a crime. A conviction under it is not merely erroneous, but is illegal and void […].’^[92] Recently, J Kennedy wrote for the US Supreme Court, in Montgomery v Louisiana,^[93] a case involving mandatory life-sentence without parole for convicted juveniles, that ‘when a new substantive rule of constitutional law controls the outcome of a case, the Constitution requires state collateral review courts to give retroactive effect to that rule.’ In Europe this is reflected in Scoppola v Italy (No 2),^[94] where the ECtHR held that ‘Article 7 para 1 of the Convention guarantees […] the principle of retrospectiveness of the more lenient criminal law.’

The void ab initio theory is tied to the robust protection of negative liberty. However, conflicts between individual rights (eg, freedom of speech and privacy), of equal legal status and protection, lack clear rules and are decided through balancing.^[95] In the age of ad hoc balancing, predominant in the jurisprudence of the ECtHR, the system of case-to-case adjudication renders the void ab initio rule virtually inoperative, to the extent that it merges into the traditional retrospective application of the judgment solely in the litigated case. But in the void ab initio theory the ‘emphasis [is] upon universal rule;’^[96] it requires clear and steadfast general rules, not malleable standards whose application depends on the totality of circumstances.

By contrast, the assimilation of the void ab initio rule with the resurrection of previous legislation is erroneous in the case of social rights. The status positivus of social rights requires legislative action to specify the corresponding legal entitlements. The fact that the cuts of pensions by the austerity measures have been found unconstitutional does not sustain the conclusion that the previous higher pensions are constitutionally mandated. The void ab initio theory preempts legislative action to regulate by subsequent legislation the preexisting pensions. Conversely, in France the Conseil Constitutionnel gave time to the legislature to pass a new law on pensions free from constitutional vice.^[97] For this reason, public law in the welfare state allows for the non-retroactivity of unconstitutionality.

Equating positive with negative rights leads to an impasse, when the public purse is drained. It is quite questionable whether the judge can invoke the previous social legislation, despite the fact that the legislature had clearly chosen to replace or amend it. The replacement of a statutory scheme can be analyzed into two distinct parts: a) the decision to supersede the legislation existing at that time, and b) the decision to introduce new regulations. Assuming the nullification of part b) as a result of the judgment of unconstitutionality, this does not render necessarily invalid part a) as well. It is for the legislature to fill the legal vacuum regarding the applicable law. In the field of social rights, the void ab initio theory raises serious questions of separation of powers, by preempting legislative power.

6 Conclusion

The void ab initio theory is a strong expression of the supremacy of the Constitution and is tied to the ideal of the rule of law. It is a theory of fidelity to pre-existing law, implemented through a strict view of legal validity; but it can also operate as a vehicle for judicial supremacy. Its privileged field of application is negative individual rights against the state, which may satisfy the best conditions for the theory: legal sanctions, on legal grounds, over adjudicative facts.

By contrast, in the welfare state, application of the void ab initio rule to social rights is problematic, because it can result into serious dysfunction and encroachments on legislative power.

The void ab initio theory, being a strong guarantee of legality, is placed under considerable pressure in a legal environment where consistency in the interpretation and application of the law is gradually but steadily eroded, because of ad hoc balancing. Of course, different systems of judicial review can influence significantly the aforementioned conclusions; but the void ab initio theory, being a strong expression of the judicial ethos of legality, has its own independent value.