non-normative propositions, which express, for example, conceptual or empirical
relationships. And, as in the context of rule application, there are different contextrelative rules for the use of evidence. And not only the rules for evidence differ
between different normative systems but also the rules for the use of the rules. Such
rules guide, for example, the interplay between different rules. If reasoning from
rules is accordingly rule-driven itself, these rules of reasoning are to be evaluated and
one can consider to what extent context-relative practices are taken into account. So,
one of the main challenges for applying the deducibility criterion is to reconstruct a
(where appropriate context-dependent) system of rules of deontic reasoning
(Alchourrón and Bulygin 1971).
4.2 The Criteria of Effectiveness and Harmlessness
The criteria of effectiveness and harmlessness are normally used in the policymaking context and the problems of using empirical evidence are relevant at this
point, because we need some evidence about the probable consequences of the uses
of the rules to apply these criteria. And normally we want to know beforehand
whether the rules fulfil the criteria. Such ex ante considerations have to use assumptions, for example concerning the behaviour of actors. So, these criteria refer to the
functions of rules and the possible consequences of rule uses. Rules are often
characterized as means or instruments. So, they can fulfil specific functions. In
general, rules can be used to evaluate and criticize actions, justify actions, coordinate
actions, learn and teach actions, predict actions and facilitate decision making.
There are some obstacles for obtaining the needed evidence for knowing whether
the criteria of effectiveness and harmlessness are fulfilled: First, we have to formulate the effects and specify them in such way that we have clear indicators or
so-called success criteria. For the problem of an operationalisation of goals a
pragmatic way is commonly used for obtaining the needed data. Furthermore, we
have to consider collateral damages. It needs to be examined what unintended or
undesired side effects are to be expected or can be observed and which side effects
are acceptable. But even if one could achieve consent about the goals or possible
indicators, there would be several ways of achieving them and the choice of the
means is a much-contested field. Who shall bear the burdens? Whose acts should be
forbidden or demanded—that is whose freedom of action should be constrained?
One of the bigger problems in this examination would be that one needs a great
deal of information for the analysis and evaluation of a rule system. The uses of rules
are guided by rules too. But there are not only different rules for using rules in
different normative systems but it is not guaranteed that these rules are applied. And
because, as Ostrom said, “[r]ules, the biophysical and material world, and the nature
of the community all jointly affect [. . .] the likely outcomes achieved” (Ostrom
2005, 16), we need information about the style of rule uses in the community in
question. Maybe that is one of the reasons why randomized controlled trials (RCTs)
cannot tell you everything you need to know: “You are told: use policies that work.
A Framework for Evaluation of Normative Solutions to Environmental Problems
117
relationships. And, as in the context of rule application, there are different contextrelative rules for the use of evidence. And not only the rules for evidence differ
between different normative systems but also the rules for the use of the rules. Such
rules guide, for example, the interplay between different rules. If reasoning from
rules is accordingly rule-driven itself, these rules of reasoning are to be evaluated and
one can consider to what extent context-relative practices are taken into account. So,
one of the main challenges for applying the deducibility criterion is to reconstruct a
(where appropriate context-dependent) system of rules of deontic reasoning
(Alchourrón and Bulygin 1971).
4.2 The Criteria of Effectiveness and Harmlessness
The criteria of effectiveness and harmlessness are normally used in the policymaking context and the problems of using empirical evidence are relevant at this
point, because we need some evidence about the probable consequences of the uses
of the rules to apply these criteria. And normally we want to know beforehand
whether the rules fulfil the criteria. Such ex ante considerations have to use assumptions, for example concerning the behaviour of actors. So, these criteria refer to the
functions of rules and the possible consequences of rule uses. Rules are often
characterized as means or instruments. So, they can fulfil specific functions. In
general, rules can be used to evaluate and criticize actions, justify actions, coordinate
actions, learn and teach actions, predict actions and facilitate decision making.
There are some obstacles for obtaining the needed evidence for knowing whether
the criteria of effectiveness and harmlessness are fulfilled: First, we have to formulate the effects and specify them in such way that we have clear indicators or
so-called success criteria. For the problem of an operationalisation of goals a
pragmatic way is commonly used for obtaining the needed data. Furthermore, we
have to consider collateral damages. It needs to be examined what unintended or
undesired side effects are to be expected or can be observed and which side effects
are acceptable. But even if one could achieve consent about the goals or possible
indicators, there would be several ways of achieving them and the choice of the
means is a much-contested field. Who shall bear the burdens? Whose acts should be
forbidden or demanded—that is whose freedom of action should be constrained?
One of the bigger problems in this examination would be that one needs a great
deal of information for the analysis and evaluation of a rule system. The uses of rules
are guided by rules too. But there are not only different rules for using rules in
different normative systems but it is not guaranteed that these rules are applied. And
because, as Ostrom said, “[r]ules, the biophysical and material world, and the nature
of the community all jointly affect [. . .] the likely outcomes achieved” (Ostrom
2005, 16), we need information about the style of rule uses in the community in
question. Maybe that is one of the reasons why randomized controlled trials (RCTs)
cannot tell you everything you need to know: “You are told: use policies that work.
A Framework for Evaluation of Normative Solutions to Environmental Problems
117