A structural approach to the distinction

For the first approach, the question is whether principles, like rules, can always be formulated in a conditional pattern, that is, one that correlates cases and solutions (this does not imply to accept — nor to reject — that norms are lin­guistic entities; it only implies acceptance of the fact that they can be expressed in some language).

The second clarification refers to which of the previous classifications of principles are relevant to the effects we are interested in now. It seems obvious that it is not the distinction between explicit and implicit principles, nor that between principles in the context of the primary systems and principles in the context of the secondary system. If there were some structural difference de­pending on whether the case regulated by a norm consists in, or refers to, the exercise of a public normative power, it would be logical to think that it would apply to principles as well as to rules. So, what we must look at is whether or not principles in the strict sense and policies have a conditional structure (that is, whether they correlate cases and normative solutions) and/ or whether the deontic operators applying to rules are the same as those applying to principles.

2.1.1. In our opinion, principles in the strict sense can always be formulated as sentences correlating cases to the normative qualification of a certain behav­iour; but that does not mean that from this perspective there is no difference to rules (and especially to action rules). The difference is that principles present the case in open form, whereas rules present it in closed form. By this we mean that, while in rules the properties constituting the case are a finite and closed set, in principles no closed list of properties can be formulated. Not only do the properties constituting the conditions of application have a larger or smaller periphery of vagueness; those conditions are not even generically determined. Thus, the kind of indeterminacy principles suffer from is more radical than that of rules (although, of course, between the two kinds of indeterminacies there can be an area of penumbra).

Robert Alexy, in developing a point which already seems to be present in Dworkin, has written that

"the decisive point for the distinction between rules and principles is that principles are norms ordering that something be implemented to the highest possible degree, relative to what is legally and materially possible. Thus, principles are mandates of optimization whose characteristic property is that they allow for different degrees of compliance, and that the amount of com­pliance they command does not depend only on the material, but also on the legal possibilities given in each case. The extent of legal possibilities is determined by principles and rules pulling into opposite directions. Rules, in contrast, are norms requiring full compliance, and insofar they can always be only complied with or violated. If a rule is valid, then it is obligatory to do exactly what it prescribes, no more and no less" (Alexy 1988,143 f.; see also Alexy 1986,75 f.).

In our view, it is true that principles can be complied with in different degrees, as far as policies or program norms are concerned, but not in the case of prin­ciples in the strict sense.

A principle like the one formulated in art. 14 of the Spanish Constitution, understood as a secondary principle, can be presented, we think, in the form of a conditional like the following: "If (condition of application) a legal organ uses its normative powers (that is, issues a norm in order to regulate a generic case, or applies a norm in order to solve an individual case, etc.), and with respect to the individual or generic case in question there is an opportunity for discrimination on grounds of birth, race, sex, religion, opinion or any other personal or social circumstance, and there is no countervailing principle of higher weight in the case at hand, then (normative solution) the organ is prohibited to discriminate on the grounds of any of the factors just men­tioned."^{1 [3]}

Here, we find the typical indeterminacy of principles only in the open­ness of the conditions of application, not in the description of the behaviour that is prohibited, i. e., that of discriminating. In some contexts, 'discrimination' can, of course, be understood as a vague term, but this kind of vagueness also exists in the standards we call 'rules'. The rule that a working woman should receive the same pay as a man differs from the above principle only in that its conditions of application are not open (thus, art. 28 of the Spanish Workers' Statute stipulates that "Employers must pay an equal salary — with respect to the basic salary as well as to salary supplements — for equal work, without any discrimination on grounds of sex"); but even here, there may be a problem of vagueness when it comes to determining whether in a particular case the condi­tions of application obtain (there may, e. g., be doubts about whether a certain activity should be considered 'work') or what the description of the prohibited behaviour amounts to (should different salary supplements for clothes on grounds of sex be considered prohibited by art.

2.1.2. When we now pass from principles in the strict sense to policies or pro­gram norms, things seem to look somewhat different. The distinctive character of policies is that this kind of standards leave their conditions of application as well as the prescribed pattern of behaviour open. As an example for a policy, we will use art. 51.1 of the Spanish Constitution, already mentioned earlier. As in art. 14, here the conditions of application are formulated in an open fashion: the provision does not state under what conditions state agencies must act as prescribed, that is, protect consumers and users. But in contrast to art. 14, art. 51 also does not command — nor prohibit — any action; instead, it commands that a purpose, a state of affairs with certain characteristics be attained, namely, that the security, health and legitimate economic interests of consumers and users are effectively protected. What actions (or courses of action) are causally appropriate for reaching this goal is not determined in the Constitution. And the objective attainment of which is commanded — that the security, health and legitimate economic interests of consumers and users are effectively protected — itself obviously is not the only goal commanded in the Constitution.

Here, we must underscore two things: on the one hand, that courses of action appropriate for attaining some constitutionally commanded purpose (e. g., making mortgages cheaper, or liberalizing the use of land, as appropriate means for making it easier for all citizens to provide decent and adequate hous­ing for themselves — an objective stipulated by art. 47 of the Spanish Constitu­tion) to the highest possible degree can have negative effects on other objec­tives also constitutionally commanded (e. g., on economic stability — art. 40 of the Spanish Constitution — in the first case, or on an "environment adequate for the development of people" — art. 45 of the Spanish Constitution — in the second); and on the other, that in many cases such constitutionally commanded objectives are interdependent: for example, the weakening of economic stabili­ty can have negative effects on access to housing or on social security systems. Thus, in contrast to what happens with principles in the strict sense, in behav­iour directed by policies the question is not which one prevails over the other in some particular case, but to formulate policies capable of attaining, to the high­est possible degree, the joint realization of all objectives. For what has been said above, in our view, Robert Alexy's theory of principles as mandates of optimization distorts the matter with respect to legal principles in the strict sense, but it seems perfectly adequate to account for policies.

Among principles (in the wide sense), policies are the pendant of end rules. The difference between end rules and policies is that the former state their conditions of application in a closed way, whereas the latter leave them open, and especially that the former stipulate an end that must be attained fully, and not only, as is the case with the latter, to the greatest possible extent, taking into account the existence of other ends and the available means.

2.2.