ISSA Proceedings 2006 – Legal Argumentation, Constitutional Interpretation, And Presumption Of Constitutionality

logo  2006Abstract: Constitutional interpretation is a very complex task. The main reason underlying this complexity is the open and abstract language of constitutional texts, mainly when it concerns their bill of rights. And when it comes to judicial review of legislation, constitutional interpretation becomes even more complex. Not only the constitution but also ordinary legislation has to be interpreted so that their compatibility can be properly analysed. Although this scheme represents common sense among constitutional scholars, the arguments used in the judicial review are the subject of fierce disputes. The aim of my paper is to analyse one of these arguments, which is frequently employed in Latin American constitutional adjudication: the presumption of constitutionality. I will argue that this kind of presumption entails many problematic issues of which constitutional scholars in Latin America are often unaware. Roughly speaking, these problematic issues can be of two types:
(1) Formal argumentation problems – concerning above all the relationship between presumption and time, as well as between presumption and proof; and
(2) Constitutional theory problems – concerning some consequences of the presumption of constitutionality in the separation of powers.

1. Introduction and definitions
In legal argumentation, presumptions often play an important role. Presuming something to be true under given circumstances – above all when it is difficult or impossible to discover the real truth – is a strategy which has been used in legal argumentation and legal decision ever since the Roman Law. Although the idea is ancient and appears, at least at first sight, quite straightforward, there is no real consensus on its precise definition and on the situations in which presumptions can be used. As will be shown further on, these two variables – definition and applicability – are of great importance to the subject of this paper, the presumption of constitutionality.
Presumptions are usually defined as the acceptance of something as true given certain conditions. But this is not enough, since it is crucial for the concept of presumption to define whether and – if it is the case – how a presumption can be defeated. In legal systems based on the Roman Law tradition, it is common to speak of two kinds of presumptions: the so-called presumptions iuris tantum and the iuris et de iure. Presumptions of the first kind can be defeated while presumptions of the latter cannot. For the aims of this paper, presumptions iuris et de iure are of no importance, since the constitutionality of an enacted law cannot be exempt from a possible defeat, at least in those countries where there is some kind of judicial review of legislation.[i] For presumptions iuris tantum, the presumed fact should be considered true unless stringent evidence to the contrary is introduced to the argumentation. In this sense, it can be said, as a preliminary working definition, that one is facing a presumption (iuris tantum) if, given certain conditions, something shall be considered true, unless there is stringent evidence to believe the contrary.
In her logic formalisation of presumptions, Edna Ullmann-Margalit (1983, 147) includes only the first part of my definition, leaving aside the reference to the evidence and to the possibility of defeat. She represents presumptions through the following formula: pres (P, Q), where P stands for the presumption-raising-fact and Q for the presumed fact. This means “P raises the presumption of Q” or “[t]here is a presumption from Q that P” (1983, 147). Nevertheless, when it comes to an explanation of the “presumption rule”, her idea is completed in the following terms: “Given that p is the case, you (= the rule subject) shall proceed as if q were true, unless or until you have (sufficient) reason to believe that q is not the case.” She calls this last part of the rule the “rebuttal clause” (1983, 149).

A more complete formalisation of the idea of presumptions can be found in Daniel Mendonca (1998, 408). According to him, the formula of presumption should take the following form: [Pro (P) & ¬Pro (¬Q)] → O Pres (Q). This means that proven that P is the case – Pro (P) – and not proven that Q is not the case – ¬Pro (¬Q) – it is then obligatory to presume Q. The importance of Mendonca’s formulation lies in its emphasis on the necessity of proving something in order to rebut a presumption. This necessity will be explored further on (see section 2.3).

In this paper, the efforts will concentrate on demonstrating two main theses:
(1) Although the concept of presumption may be fairly straightforward in many legal subjects, its applicability within the constitutional argumentation – under the label “presumption of constitutionality” – entails several formal problems, above all those concerned with the relationship between presumption and time, and between presumption and proof, as well as between presumption and conditions.
(2) The presumption of constitutionality, when allied to other canons of constitutional interpretation, may have – and often has – paramount consequences for the separation of powers and for the role of judges in the judicial review of legislation (it will be shown that the use of the topos “presumption of constitutionality” is the first step to a judicial activism “disguised” as judicial restraint).[ii] Although it is intuitive to think that the contrary is the case, i.e., that presuming the constitutionality of an enactment of the legislative power is both a respectful approach and an exercise of judicial restraint (in this sense, Stokes 2003: 345 ff.), it will be shown that judges often use this kind of presumption and this alleged respect as a excuse to correct, change or extend the textual meaning of a statute.

2. The presumption of constitutionality
In a past decision of the Brazilian Supreme Court, Justice Moreira Alves argued that when interpreting a statute the Court must presuppose its constitutionality. According to him, this should be always the working hypothesis from which the court should begin.[iii] This statement can be understood in at least two different ways. On the one hand, it can be said that it is a plain triviality, since it would be a nonsense to think that legislators act always unconstitutionally and that it is the judges’ task to demonstrate the contrary. On the other hand, it can be understood as a presumption that can be rebutted in some cases. In this case, although it is not possible to speak of a triviality or of nonsense, resorting to the idea of presumption is not unproblematic.
In order to demonstrate this, it is first of all necessary to analyse three issues that undermine the possibility (or the usefulness) of the presumption of constitutionality in any sense. The first one is related to the concept of time, the second to the concept of condition, and the third to the concept of proof.

2.1. Presumption and time
The first argument against the possibility of a presumption of constitutionality that should be discussed is related to some problematic issues concerning the relationship between presumption and time. The presumption of constitutionality can be understood as the presumption that, whenever the legislator enacts a statute, he always intends to act in accordance with the constitution. But this idea considerably weakens the expected argumentative strength of the presumption of constitutionality, for it is only possible to presume that the legislator respects the constitution that was in force at the time the statute has been enacted.[iv] Hence, the presumption of constitutionality could hold (if at all) in the Brazilian case only for laws enacted after October 5th, 1988, which is the date of the promulgation of the constitution presently in force. And since this constitution has been amended 52 times since its promulgation, it is allowable to suppose that, if a statute apparently contradicts an article that has been changed, then the presumption can only hold if the statute has been enacted after this constitutional change.

2.2. Presumption and conditions
It has been shown that the presumption formula entails not only the fact to be presumed, but also the conditions under which this same fact is to be presumed, i.e. something is to be presumed as true under certain conditions or circumstances. For example: many civil codes stipulate that a child born at least 180 days after wedlock is presumably legitimate; or that if husband and wife die in the same car (or plane, or train) accident, it should be presumed that the deaths were simultaneous.
However, this model is impossible to follow when it comes to a presumption of constitutionality, for it is impossible to define under which conditions or under which circumstances a statute should be presumed constitutional and under which conditions it should not. Indeed, if there were any reason to believe in a presumption of constitutionality, one should believe in it in every case. The “given-clause” is totally absent. There is no “given the child was born at least 180 days after wedlock …” (and not 179 day or less) or “given husband and wife were in the same plane that crashed …” (and not in different planes), but only “given that a law is enacted”, which states no real condition or circumstance.
Hence, to state that a law is always presumably constitutional is the same as saying that every law is constitutional unless someone (a constitutional court, for instance) declares otherwise. But this is not really a presumption, since presumptions start with the statement of some conditions, as already shown. The examples above demonstrate the idea. Although it would be possible to imagine a norm stating that every child is legitimate until a judges decides otherwise, this norm would not express any kind of presumption. Actually, it could be said that such a norm would be completely superfluous, since it would be nonsense to state the contrary (“every child should be considered illegitimate until the contrary is proved”).

2.3. Presumption and proof
When it comes to a legal presumption, as already shown, it is necessary to accept that something is true if, given certain conditions, there is no proof to the contrary. Following this pattern, art. 1597 (1) of the Brazilian Civil Code, which states that a child born at least 180 days after wedlock is presumably legitimate, stipulates a presumption, as already stated.
The traditional idea of legal presumption (iuris tantum), already outlined in this paper, presupposes the possibility of demonstrating the contrary of what the presumption stipulates, i.e. it presupposes the possibility of rebutting the presumption when it is possible to prove that the presumed fact is not true. In the case of the example mentioned above, it is possible, through a DNA test, to prove that a child is not legitimate, even if he or she were born more than 180 days after wedlock. But this kind of rebuttal is impossible when it comes to a presumption of constitutionality, for the simple reason that constitutionality and unconstitutionality are not subject to proof. “Being” constitutional or unconstitutional are not inherent features of laws. Contrary to what common sense seems to propose, the process of constitutional review of legislation is not a kind of search for a “genetic code” – congenital to the enacted law – waiting to be discovered by legal scientists.
Therefore, when legal scholars and legal practitioners argue that a statute cannot be declared unconstitutional unless it is provably unconstitutional, they are mistakenly transposing the idea of a factual presumption to an argumentative presumption without being aware that this latter kind of presumption (argumentative presumption) cannot be confirmed or rebutted according to the same rationale, for the constitutionality or the unconstitutionality of a statute, despite being subject to legal argumentation, is not subject to any kind of proof or evidence.

3. Constitutional Theory and Separation of Rights
Despite the various theoretical inconsistencies which surround it, the presumption of constitutionality has been frequently used by courts in many countries. Its problematic consequences, however, for both the constitutional review of legislation and the separation of powers often remain unnoticed. To understand the kind of consequences I am referring to, we could take a brief look at some judicial decisions by courts in three different countries. In the United States, the Supreme Court (and also other state courts) often resorts to the so-called “constitutional avoidance canon”, in the following terms:
“Under this canon of statutory construction, the elementary rule is that every reasonable construction must be resorted to in order to save a statute from unconstitutionality” and “as between two possible interpretations of a statute, by one of which it would be unconstitutional and by the other valid, our plain duty is to adopt that which will save the Act.”[v]

The Federal Constitutional Court of Germany has several decisions with a similar view on the question, the leading case in this matter being the following decision:
“[a] statute should not be declared void if it is possible to interpret it in a way compatible with the constitution, for it is necessary not only to presuppose that a statute is compatible with the constitution, but also that this presupposition expresses a principle, according to which, in case of doubt, a statute should be interpreted in accordance to the constitution.”[vi]

And an example from the Brazilian Supreme Court:
“[The] interpretation of the assailed statute should start from a working hypothesis – the so-called presumption of constitutionality – from which derives the rule that between two possible understandings about the assailed norm, the one that is in accordance with the constitution should prevail.[vii]

To put in a nutshell: according to the views expressed in the transcribed decisions, among all the possibilities of interpreting a statute, judges should always prefer the one that sustains its constitutionality. According to this view, this would confirm the presumption of constitutionality. Additionally, by acting in this way courts would respect the separation of powers as well as the work of the legislator.

I argue that this pattern of argumentation not only has several theoretical flaws – as demonstrated in sections 2.1, 2.2, and 2.3 above – but also that it means neither a preservation of an equilibrium between courts and legislators, nor a deference to the work of the democratic legislator. On the contrary: in the way the presumption of constitutionality is usually applied, it grants much more power to judges than they already have without it. In order to demonstrate this thesis, I will call upon the so-called “interpretation in accordance to the constitution” (Verfassungskonforme Auslegung) mentioned in the decision of the German Constitutional Court, as transcribed above. This canon of interpretation – widely accepted in other European[viii] as well in Latin American countries[ix] – states simply that judges have the duty to prefer the interpretation of a statute that maintains its constitutionality (interpretation in accordance with the constitution). In other words, judges are obliged to try to save the statute from unconstitutionality and hence to save also the presumption of constitutionality from any kind of rebuttal.

4. Correcting the meaning of the law
I argue that, in these cases, “respect for the legislator” is merely a commonplace. The court actually gives its own interpretation of the statute, in order to make it compatible with what the same court – and nobody else – thinks is constitutional. At this point, one can ask: But is this not exactly the task of a constitutional judge? It is indeed. The task of constitutional judges is exactly to interpret a statute in order to check its compatibility with the constitution. However, for this task – usually assigned by the constitution itself -, there would be no need for resorting to concepts like “presumption of constitutionality” and “interpretation in accordance with the constitution”. But if not, then why do judges do it so frequently?
Judges – including constitutional judges – normally feel uncomfortable with the idea of “creating the law”. They usually regard their task as a merely interpretative task. To justify such a view, the Brazilian Supreme Court uses the Kelsenian dichotomy between “positive legislator” and “negative legislator” (Kelsen 1929, 34-35). According to Kelsen, a constitutional court can only act as a negative legislator, i.e. the court can, at the utmost, annul a statute because of non-conformity with the constitution. But a constitutional court cannot create norms positively. However, the Brazilian constitution – like many other European and Latin American post-war constitutions – poses, by raising a very large array of themes to the constitutional level, new challenges to constitutional judges. To face these challenges, the judges need more than the simple dichotomy between negative and positive legislator. But for those judges who are unwilling to abandon the Kelsenian dichotomy and still pretend that constitutional judges are “no more than the mouth that pronounces the words of the law” (Montesquieu 1748, XI/6), the presumption of constitutionality and its main consequence – the duty to save the enacted law by interpreting it “in accordance with the constitution” – can be very useful. By resorting to this kind of argumentation, they can still – at least apparently – remain faithful to the “negative legislator dogma” and, at the same time, correct or extend the work of the legislator whenever they consider it convenient.
Except in unimportant and trivial cases, this occurs because the duty to save the law from unconstitutionality implies a possibility – and frequently a necessity – of altering its meaning, especially when saving the enacted law implies going beyond what the legal text prescribes. I am of course not unaware of the fact that interpreting the law is always ascribing a meaning to the law, a meaning that may not be the same meaning the parliament majority had in mind when it passed it. This is, per se, not a problematic issue – except for those who believe that interpreting the law is to search for the legislator’s intent. But what is problematic is to mask this fact as an alleged “deference to the legislator” and behind a unjustified and theoretical unsound presumption of constitutionality.

[i] In this sense, it can be said that a presumption of constitutionality would be iuris et de iure in those countries where parliament is sovereign.
[ii] I do not intend to take sides in the dispute between activism and restraint, and this would be in any case not required for the aims of this paper. As is clear in the text, the problem is not activism as such, but a “disguised activism”, in which judges pretend to exercise restraint while modifying the meaning of enacted law.
[iii] Rep. 1417 (1987). See RTJ 126, 48 (53).
[iv] In this same sense, see, for the German case, Skouris (1973, 98), Gusy (1985, 218) and Bogs (1966, 22). For the case of the United States – by Scheef (2003, 530 ff.).
[v] Rust v. Sullivan, 500 U.S. 173 (1991) – emphasis added. This is a very old canon within the US Supreme Court. See for instance Hooper v. California, 155 U.S. 648 (1895). More recently – and with details about the “constitutional avoidance canon” – see Clark v. Martinez, 543 US 371 (2005). See also Vermeule (1997, 1949).
[vi] BVerfGE 2, 266 (282) – emphasis added.
[vii] RTJ 126, 48 (53) – emphasis added.
[viii] See, for instance, the following decisions: Portuguese Constitutional Court – decisions 327/99 e 466/00; Italian Constitutional Court – decisions 138/1998 and 139/1998; ; Austrian Constitutional Court – decision 11.576/1987.
[ix] See, for instance, the following decisions: Columbian Constitutional Court – decisions C-496/94 e C-109/95; Chilean Constitutional Court, decision 309/2000; Brazilian Supreme Court – decisions RTJ 173, 424; RTJ 181, 54; RTJ 167, 376; RTJ 178, 919; and RTJ 167, 363.

Bogs, Harald (1966). Die Verfassungskonforme Auslegung von Gesetzen. Stuttgart: Kohlhammer.
Gusy, Christoph (1985). Parlamentarischer Gesetzgeber und Bundesverfassungsgericht. Berlin: Duncker & Humblot.
Kelsen, Hans (1929). “Wesen und Entwicklung der Staatsgerichtsbarkeit”, VVDStRL 5: 30-88.
Mendonca, Daniel (1998). “Presumptions”, Ratio Juris 11 (4): 399-412.
Scheef, Robert W. (2003). “Temporal Dynamics in Statutory Interpretation: Courts, Congress, and the Canon of Constitutional Avoidance”. University of Pittsburgh Law Review 64: 529-587.
Skouris, Wassilios (1973). Teilnichtigkeit von Gesetzen. Berlin: Duncker & Humblot.
Stokes, Michael L. (2003). “Judicial Restraint and the Presumption of Constitutionality”, University of Toledo Law Review 35: 347-375.
Ullmann-Margalit, Edna (1983). “On Presumption”, Journal of Philosophy 80 (3): 143-163.
Vermeule, Adrian (1997). “Saving Constructions”, Georgetown Law Journal 85: 1945-1977.