1 Topic and Structure of the Article

The aim of this article is to find rules, criteria for good public political argumentation, by which political measures in the broad sense can be justified. There are several layers of such rules, from linguistic to process rules. This article deals with epistemic rules for arguments (that describe the structure, in part the content as well as the conditions of their usage), whose observance should guarantee that the thesis of an argument is acceptable and can also be recognised as acceptable by an addressee. I will proceed as follows: 2. On the basis of elements of a theory of deliberative democracy, I will try to determine how political measures need to be justified and try to specify what kinds of theses need to be justified in doing so. 3. Next, in addition to the central thesis (‘m is … morally optimal’), other important but subordinate theses are determined that must be justified in public argumentation. 4. A brief discussion of various approaches in argumentation theory filters out the approach that is most likely to provide criteria for epistemically good arguments by which such theses can be justified: it is the epistemological approach. 5. Relying on what has been elaborated, epistemological criteria for good arguments are then presented by which theses of the aforementioned kinds can be justified. 6. Finally, an analysis of factual political argumentation shows to what extent such standards are also empirically followed and how the factual argumentation quality can be improved.

2 Determining the Central Thesis of Public Argumentation via its Function in Deliberative Democracy: Serving the General Good

Public arguments, as they are understood here, are political arguments about politically enforceable measures. The locus of these political arguments is deliberation, especially in political institutions and in the broader political public, as it is characteristic of deliberative democracies. Advancing these arguments has a specific function in deliberative democracy. In the following, the epistemic norms for public arguments will be determined in two broad steps: 1. What kinds of (central) theses must be justified in a deliberative democracy in order to fulfil the function of public argumentation for deliberative democracy? 2. How can such theses be justified?

In order to be able to conduct the discussion on the function of argumentation in deliberative democracy on a better basis, what exactly is meant here by ‘deliberative democracy’ should first be definded.

  • DDD Definition of ‘deliberative democracy’: ‘Deliberative democracy’ for mass societies means: 1. representative, parliamentary democracy, 2. in which the official, institutional political decisions are flanked by lively political debates with broad participation in a critical public sphere (in Habermas' sense) outside the political institutions, in which upcoming and completed political decisions are constantly critically prepared and commented on; 3. the epistemic hallmark of deliberative democracy is that it is characterized by intensive and extensive argumentation in the narrow, epistemological sense (with effective truth orientation), which thus aims to establish the truth about the issues under discussion on the basis of epistemologically founded criteria, and that this argumentation is also encouraged (epistemic orientation).Footnote 1

Most Western democracies have at least some deliberative elements in this sense.

Deliberative democracy combines two functional goals: an epistemic goal, namely, to find, in particular with the help of the deliberative component, truths important for the decision or even the one decisive truth, and a participatory, republican goal, namely, to give, with the help of the democratic component, all members of the community the same power to determine the policy of this community (cf. e.g. Goodin 2000). While these two components can be clearly separated externally—argumentative discourse versus voting—they seem to belong to completely different and mutually exclusive logics of decision making: decision by finding the truth versus decision by majority and thus social power. With respect to the truth-seeking component, realistically speaking, the participation of the demos is more of a hindrance to finding the truth; an expert discussion would be much more appropriate. Furthermore, it is not clear what truth has to be sought after, what thesis has to be argued for for promoting a certain political proposal or for providing its definite justification. And with respect to the participatory component, the deliberation and general discussion is superfluous for a decision by majority. Attempts by defenders of deliberative democracy to reduce this tension and to connect both components organically (e.g. Habermas 1983; 1994; Cohen 1986; 1989; 1996; 2009; Estlund 2008; Mansbridge et al. 2010) are in my opinion not convincing so far.

How can the tension between these two components be resolved? One attempt to resolve it is the liberal conception of democracy. According to this conception, the goal of citizens in a democracy is to advance as many of their own interests as possible. The complete realisation of one's own interests is, of course, an unrealistic goal; but a realistic goal is to achieve as many of one's own interests as possible, especially by forming coalitions that guarantee that the successful coalition members will have a particularly large number of their own interests fulfilled. Deliberation takes on a double function here. For one thing, it serves to clarify all possible epistemic issues so that the final decision is as rational as possible; but the other function is not epistemic at all; rather, deliberation is a negotiation to secure a majority for a particular proposal through compromise.—But this is not what happens in politics. Rather, the various participants in the debate try to show that their own proposal is most conducive to the general good. One reason why this is so is that in a complex democracy with a plethora of completely disparate interests of the most diverse groups, no sufficiently strong agreement can be found with the liberal model. Another reason is that the liberal recourse to the interests of a number of those affected does not ensure a moral result: certain minorities of all kinds are neglected; a coalition of the majority can thus exploit, suppress or even try to eliminate a very large minority. At least a significant proportion of citizens does not want this, but would like the state in general and individual political measures in particular to follow moral principles (Mansbridge 1990). This actual behaviour of most citizens has puzzled advocates of a liberal market model of democracy because they assume an image of citizens as rational egoists who seek to maximise the fulfilment of their own interests. But the tendency to make moral, common good or ideological decisions in politics and beyond is factually very strong (ibid.).Footnote 2

These empirical findings mean that in deliberative democracies, at least to a considerable extent, a moral-instrumentalist conception of the state in particular and of social institutions in general is realised: these institutions serve to promote (in political philosophical terminology) the general good or to realise (in moral philosophical terminology) moral values—which includes preventing corresponding deteriorations. They are in the main tendency created by people to promote the common good in a new way, reformed to promote it better; and the actions of public social institutions in the narrow sense (i.e. institutions capable of action) aim at creating a new common good. The following normative considerations build on this empirical analysis and are intended for people who embrace this objective and the moral-instrumentalist conception of social institutions on moral grounds and out of concern for the general good.Footnote 3

According to this conception, the epistemic and participatory components of deliberative democracy have, in the first place, the following functions: 1. The epistemic goal of deliberation is to determine those policies that most promote the public good. 2. The participatory component is needed for the following reason: The discourse only provides knowledge, not its implementation. It is the democratic participation of all that gives moral insights their power in the first place—to the extent that individuals are guided by these insights.

Both the participatory and the epistemic component, however, have a further function: So far, neither among ethicists nor in the general public is there consensus and there are very different ideas on how the public good should be defined (Rawls 1989). Nor is it to be expected that agreement will be reached on this in the near future. (Over time, however, the respective discussions marginalise the worst justified ethics; fewer and fewer people adhere to them. And newly introduced ethical systems, in addition to formal and justification advantages, often have a compromise character, or they expand the scope of moral commitment. All this thus leads to a certain convergence of the positions actually held among ethicists. The general public and political discourses can benefit from these theoretical developments if they remain linked to these theoretical discourses through the critical public sphere, education of citizens and representatives, and also have an impact on the theoretical discourses. This is the particular moral function of the epistemic elements of deliberative democracy. In short, in the political public sphere of a deliberative democracy not only applied-ethical but also normative-ethical (and meta-ethical) discussions are held.) 3. Public deliberation then also has the function of advancing the discussion on the criteria of the public good in substantive terms. 4. And the democratic vote also has the function, as long as no epistemic agreement has been reached, of deciding on the criteria to be applied for determining the public good in particular decision situations according to the strongest social support. 5. Another essential function of the participatory component in determining the content of morality regards the conflict in principle between the prudential and the moral goals of individuals: To what extent should these goals be realised? How many resources do individuals want to invest in one and the other? Through their participation, citizens also determine collectively how much they are willing to give for the realisation of moral goals and thus determine the size of the budget for moral tasks. (Aside: Act utilitarianism completely ignores this question and assumes that the moral share of the total individual budgets must be 100%.)—The deliberative component of democracy could be limited to parliamentary debates. The extension of the deliberative component to what is now understood in a narrower sense as “deliberative democracy”, i.e. in particular the strong participation of a critical public sphere (see the above definition of ‘deliberative democracy’), has additional functions: 6. More people and their specific knowledge are included in the determination of the general good. This broadens the information base for determining the general good. 7. Even more weight is given to the interests of all. In addition, it gives more weight to those conceptions of the general good according to which the well-being of each individual must enter into the determination of the general good, namely the welfare-ethical conceptions of the general good. 8. The participation of individuals in the debate also improves their political knowledge, which is reflected not only in their contributions to the debate, but also in their voting behaviour, with which they demand morally better justified measures.

3 What Other Important Types of Theses Need to be Defended in Public Arguments?—The final List of (More Precise) Theses

On the basis of the approach just developed, I would now like to determine more precisely what kinds of theses actually need to be argued for mainly in public discourse.

If public discourse also takes part in the normative-ethical discussions about the correct moral principles and even in the metaethical, then there are (at least) three layers of public argumentation:

  1. 1.

    the meta-ethical and normative-ethical level of the discussion on the criteria of the common welfare and the moral good and the right in general,

  2. 2.

    the applied-ethical level of discussion on the moral value (according to these criteria) of proposed public measures—'This measure promotes the public good to that and that extent'—or their obligatory nature—'Our obligation to do A requires the public measure p.' There are then various moral desirability criteria: utilitarianism, prioritarianism etc. as well as formally valid and informally more or less valid social norms; and

  3. 3.

    the level of empirical subquestions concerning moral value, above all the morally relevant consequences of public measures, but also e.g. about the possible social acceptance or support of some measure under discussion.

What has been said about the applied-ethical level needs further explanation: The political community makes moral claims or desiderata socially binding by publicly proclaiming them and providing them with threats of—formal, i.e. legal, or informal—sanctions. More specifically, this involves several categories of actions and measures: 1. deontic measures that are or could be the consequence of an existing formal, i.e. also legal, moral duty or of an informal, but nonetheless socially more or less valid duty, namely, 1.1. actions or measures of fulfilling a duty and 1.2. of supporting the respective norm by controlling its observance, applying sanctions etc.; and, 2. moral-political measures for the moral improvement of the system of norms and institutions: Bad norms and institutions are abolished or replaced by new, better ones; new norms and institutions that extend moral commitment are introduced.

So, according to the outlined conception, which theses mainly have to be justified in public argumentation? And what types of arguments can be used for this?


  • T1: Assessment of the individual utility of each person affected: Utility assessments or evaluations of possible public measures from the point of view of individual affected persons or groups of affected persons are prudential value judgements—these can be justified with the help of practical arguments for prudential value judgements (Lumer 2014).

  • T2: Moral evaluation: aggregation of individual utilities: The aggregation of individual utilities in a moral or social evaluation is based on one of the individual welfare ethical definitions of 'moral value' like the utilitarian definition (utilitarian moral value of p = the sum of the individual prudential utilities of p for all the single sentient beings) or a prioritarian definition (prioritarian moral value of p = the sum of the individual prioritaristically weighted prudential utilities of p for all the single sentient beings (Lumer 2005b)) or egalitarian definition (egalitarian moral value of p = the sum of the individual prudential utilities of p for all the single sentient beings minus the measure of the inequalitiy of the distribution of these prudential utilities (Sen < 1973 > 1997)) etc.; these definitions provide the mathematical formula for aggregation—this mathematical formula plus the information on the various individual utilities are the main premises of mathematical, i.e. deductive arguments for moral value judgements (see below Sect. 5.2; Lumer 2005a).

  • T3: Consequences and technical feasibility of specific public measures: On the epistemically strongest level, theses on the feasibility and consequences of measures are scientific forecasts of all kinds—which can be sustained often by deductive, but mostly by probabilistic arguments (Lumer 2005a; 2011).

  • T4: Social passability and enforceability of particular measures: Theses on the enforceability of measures rely, again at the epistemically strongest level, on social science findings on widespread attitudes and are predictions about social behaviour—which can be substantiated by statistical and probabilistic arguments (Lumer 2011).

  • T5: Legality and moral permissiveness: That a discussed measure complies with or is required by the currently valid social norms is a legal (or deontic) thesis. In the simplest case such theses can be derived from the factual description of the situation and the text of the norm. In more complicated cases interpretations are necessary because of vagueness or contradictory norms. Such interpretations are the subject of long legal debates. The most important hermeneutical proposals thereon are: recourse to the intention of the legislator and moral optimisation within the scope of vagueness—So, in simple cases deductive arguments are sufficient, whereas in more complex cases interpretive arguments for establishing the legislator's intention or practical arguments for prudential value judgments in connection with deductive aggregation for establishing the morally best interpretation have to be added (Lumer 1990; 1992; 2005a; 2010).

  • T6: Meta-ethical and normative-ethical discussion: The ultimate thesis of a meta-ethical and normative-ethical discussion about public measures is that a certain moral criterion is adequate, correct etc. How exactly to argue in this case is very controversial in ethics. Here I list only a few types of arguments which must be used in any case for this purpose—deductive arguments, practical arguments, interpretive arguments (Lumer 1990: 180–209; 221–246; 320–358; 2005a: 221; 2014; 2010; a proposal for a specific strategy for justifying moral principles plus the determination of the various arguments to be used is provided in: Lumer 2015).

  • T7: Secondary justifications: reference to experts: A number of the arguments listed so far are scientifically and epistemically inaccessible to the general public involved in the public debate, but often also to political experts. The best, i.e. generally understandable and maximally valid, substitute for these primary arguments is then the recourse to experts who provide some guarantee for a specific thesis—this kind of justification is provided in arguments from expert opinion (Lumer 2020). However, such arguments must also be able to deal with the complex case of experts contradicting each other (ibid.).

  • T8: Best strategy for politically pushing through a measure: In discussions of moral strategy, the question is not only whether and how a certain political measure can be socially pushed through at all, but what is the (morally) best strategy for pushing it through. This is a question of moral evaluation, for which the arguments already discussed above (T2) can be used.

These are only the central theses and arguments of the discourse. All reasons put forward for them or in them can, if necessary, in turn be justified argumentatively, and so on, whereby more and more new types of theses and arguments can then be introduced into the political discourse in ever more branching arguments.

4 The Epistemologically Appropriate Argumentation-Theoretical Approach

We now have to discuss which argumentation-theoretical approach provides criteria for good arguments, which do justice to the function of argumentation as described at the beginning, namely, for one, to guarantee the truth or at least the epistemic acceptability of the correctly justified thesis—because only in this way can we find out the really good options—and for another, to also make this truth or acceptability recognisable to the addressees of the argument. I briefly discuss here Pragma-Dialectics, Walton's argument schemes, criteria of Informal Logic and of epistemological argumentation theory. Of course, these cannot be detailed discussions here, but only relatively superficial considerations as to whether the respective theory contains criteria of the desired kind for arguments for the theses mentioned at all and whether the approach has at least epistemic aspirations or, better, (in principle) is able to provide the epistemically required.

Pragma-Dialectics (Eemeren & Grootendorst 1984, 2004) focuses mainly on the procedural rules of discourse and leaves the selection of criteria for good arguments to the agreement between the participants in the opening phase of the discussion. This approach is not very helpful for our purposes because, if we are to engage with this framework, we would like to know what would be a good agreement for our purposes. Occasionally, Pragma-Dialectics does provide hints for good arguments on the "product level", but these are strongly oriented towards deductivism. This is not very plausible as such, because it does not do justice to the existence of weaker than deductive arguments. Deductivism does not fit at all with the probabilistic (T3, T4), statistical (T3, T4), interpretive (T5, T6) and practical arguments (T1, T6) as well as arguments from expert opinion (T7) required for the respective theses.

Walton's argumentation schemes approach (Walton 1996; Walton et al. 2008; Walton & Sartor 2013) develops many, in its most extensive version (Walton et al. 2008)) 104, argumentation schemes; but it does not provide schemes for probabilistic (T3, T4) and statistical arguments (T4), however, for practical arguments. Even so, these are mostly quite simple schemes for ends-means arguments, which therefore do not address the actually complex side of practical arguments, i.e. weighing up various advantages and disadvantages as well as a final quantification of the resulting utilities. The schemes also do not provide any approach to the ethical weighing of the interests of different persons (T2). Moreover, Walton's schemes—his (ex post) justification attempt (Walton & Sartor 2013) notwithstanding—are descriptions of empirically found argument patterns but have not been checked or even constructed epistemologically with respect to their capacity to guarantee the thesis' acceptability; in fact many of Walton's schemes endorse ineffective arguments, which too rarely lead to acceptable theses (Lumer 2022a).

Informal Logic (e.g. Blair 2012 (in particular 33–47); Govier < 1985 > 2010: chs. 3; 7–11; 1992; Johnson 2000: 189–209) has legitimately insisted on rational criteria for inferences and the acceptability of premises, but has dealt with these issues only in relatively general terms (the criteria are: relevance and acceptability of the premises as well as their sufficiency [?] to justify the conclusion), has not provided a theory of how to develop such criteria, and accordingly has elaborated only few precise criteria for good arguments. Among Informal Logicians, Trudy Govier has still developed most of such criteria (< 1985 > 2010). But even here, probabilistic (T3, T4), interpretive (T5) and practical arguments (T1, T2) are not treated. And under the title "conductive arguments" she mixes e.g. practical arguments, which list advantages and disadvantages of an option, with arguments from sign, where there is circumstantial evidence for and against a certain empirical proposition.

Within the epistemological approach (Biro & Siegel 2006; Feldman < 1993 > 2013; 1994; Goldman 1999: 131–160; Lumer 1990; 2005a; Siegel & Biro 1997), instead, there are theories of argumentation that fulfil the required conditions. In the Practical Argumentation Theory, precise criteria for good arguments of various types are developed. These criteria are each divided into two groups: criteria for the argumentative validity of the argument and criteria for the adequate use of the argument for epistemic purposes. The criteria for argumentative validity are based on so-called epistemological principles, which are again taken from epistemology in the broad sense, i.e. also from probability theory or rational decision theory and philosophical desirability theory. The content of these principles is that they list sufficient conditions for the acceptability of a thesis. Good arguments must then go through these conditions in detail and make true and verifiable claims about the way in which the conditions are fulfilled for the thesis in question. If the conditions of the epistemological principle are fulfilled for the respective thesis, then the thesis is true or at least acceptable. That the argument is also adequate for cognitive purposes implies above all that it is epistemically accessible, i.e. that the subjects of cognition, the addressees, can verify the assertions made in the argument’s premises, so that—in the case of a positive result—they can also subsequently note that all the conditions of the epistemological principle are fulfilled, whereby they have then recognised that the thesis in question is true or acceptable. (Lumer 1990; 2005a) This theory is also elaborated to such an extent that it contains criteria for good arguments for all types of arguments required in the above (Sect. 3) list. I have already anticipated a proof of this assertion by adding to the entries in the list of the most important argument types needed in public argumentation respective references in each case. With all this, the epistemological approach thus provides what was sought here: precise criteria for the necessary types of arguments, whereby adherence to the criteria for argumentative validity guarantees the acceptability of the thesis due to their epistemological foundation, and the additional adherence to the criteria for situational adequacy enables the addressee to recognise the acceptability of the thesis.

The result of this brief discussion is thus quite unambiguous: The only theory of argumentation that provides the desired criteria for good arguments is the epistemological theory of argumentation and therein in particular the Practical Theory of Argumentation with its precise criteria for various argument types. With the help of such criteria for good arguments, it is then possible to justify a rational policy whose goal is really the general good, instead of, for example, serving populist prejudices that are widespread in political discourse. And it is possible to make the reasons for this rational policy comprehensible to fellow citizens.

5 Rules for the Main Types of Arguments Needed in Public Argumentation

I cannot discuss all the sets of criteria for the types of arguments mentioned in the list of Sect. 3 in detail here. For that, please refer to the bibliographical references. Instead, I would like to present here the criteria for two types of arguments that are actually central to political discourse, but which are rather neglected both in practice and in argumentation theory: prudential practical arguments for theses about the individual utility of specific measures (T1) and moral practical arguments for the moral value of such measures (T2).

5.1 Criteria for Epistemically Good Arguments for Prudential Value Judgements (T1)

The basis of prudential practical arguments, the epistemological principle that guides them, is the definition of ‘expected utility’. The expected utility of an event p for a subject s is the sum of the products of probability and utility for s of the various (relevant) consequences of p (plus the intrinsic utility of p for s). More formally:

  • Definition of ‘expected utility’:

  • An event/state p has for a subject s the expected utilityexpected utility u :=

  • 1. let c1,..., cn be all possible consequences or implications of the event p that are relevant for s and do not overlap;

  • 2. the probabilities that p really produces these consequences are l1,..., ln ("l" stands for " likeliness");

  • 3. the utilities of c1,..., cn for s are u1,..., un;

  • 4. the intrinsic utility of p is up;

    then the expected utility of p for s is: u := l1u1 + … + lnun + up = up + Σi liui.


(Elementary) prudential practical arguments are then arguments for value judgements that an event p has the (expected) utility u for a subject s; and their arguments consist of judgements that follow the conditions of the expected utility definition and specify how these conditions are fulfilled in the concrete case. In the following conditions for epistemically good prudential practical arguments, first the ideal structure of these arguments is described (PA1); the thesis is the value judgement to be justified (PA1.1); the reasons address the individual conditions of the definition of expected utility (PA1.2). This is followed by the condition for the argumentative validity of such structures (PA2). Ideal practical arguments are cumbersome and long; a further condition therefore allows for simplifications that do not affect the validity of the argument (PA4). The last condition specifies when a valid practical argument is epistemically accessible to someone and thus also adequate to guide them in recognising the thesis (PA5).


  • Conditions for an epistemically good prudential practical argument:


  • I. Argument structure and argumentative validity:

  • PA1: Components and structure of prudential practical arguments:

  • PA1.1: Thesis: ‘The event p has utility u for the subject s.’

  • PA1.2: The reasons:

  • PA1.2.1: n judgements of the form: ‘The probability that event p has consequence / implication ci is li.’ (li>0);

  • PA1.2.2: for each of the reasons from PA1.2.1, a reason of the form: ‘The utility of consequence / implication ci for subject s is ui.’;

  • PA1.2.3: a judgement of the form: ‘The implications c1, ..., cn are all consequences of p that are relevant to s—where ‘relevant’ means: the product of the probability li and the utility ui is greater than the relevance threshold r.’;

  • PA1.2.4: a judgement of the form: ‘The intrinsic desirability of p is up.’;

  • PA1.2.5: finally, a judgement of the form: ‘The sum of (p's intrinsic desirability) up and of all products liui is u.’ (l1u1 +··· + lnun + up = u.)

  • PA2: Argumentative validity: Truth guarantee: All reasons from PA1.2 are true.

  • PA4: Liberalisation: Simplification rules: All of these reasons must be true, but not all of them must be made explicit; this applies in particular to PA1.2.3, PA1.2.4 if p has no intrinsic desirability, P1.2.5.

  • II. Situational adequacy:

  • PA5: Situational adequacy: A valid argument with properties PA0-PA4 is adequate to convince a hearer h of the thesis if the following conditions hold:

  • PA5.4: Consequence knowledge: The addressee h already believes in the consequence relations from PA1.2.1 or can recognise them immediately.

  • PA5.5: Value knowledge: The addressee h already believes in or can immediately recognise the desirability judgements from PA1.2.2 and PA1.2.4.

  • (Abridged from: Lumer 2014; cf. also: Lumer 1990: 319-366.)

In many simple cases, above all when there are only one or two relevant consequences, the utilities can also be indicated without numbers by vague indications of magnitude such as: ‘very good’, ‘small disadvantage’.—The reasons PA1.2.2 with the evaluation of consequences are prima facie rather strange, because they have the same form as the thesis justified with the argument and thus seem to lead to a vicious circle. But the ideas behind them are: Firstly, the complex desirability of p is already broken down to the desirability of the various consequences ci of p, which can then be estimated more easily. Secondly, the desirability of the consequences can possibly be justified again via the desirability of the consequences of the consequences of p until one arrives at intrinsically desirable consequences. There is a long philosophical debate about what is intrinsically desirable or relevant. The main theory in this debate is rational hedonism, that is: precisely one's own feelings are intrinsically relevant to oneself; and their desirability is identical to their (positively or negatively directed) intensity times their duration. This theory most likely is not entirely correct; but it is probably the best approximation to the correct theory. If this is so—and for the sake of simplicity here I assume it is—then one can simplify the justification of individual utility judgments by using QALYs (= quality-adjusted life years) as utility units and by relying on empirical studies from happiness research on the influence of various factors on affective well-being in determining the hedonistic quality of a life in a given time period. One QALY corresponds to the average affective well-being in a reference population over a period of one year. Someone who experiences two years with the average affective well-being of society thus has a utility of 2 QALYs in this time. Someone whose average affective well-being in one year is only half that of the average of society has a utility of only 0.5 QALYs in that year.

5.2 Criteria for Epistemically Good Arguments for Moral Value Judgements (T2)

I have treated the arguments for theses on the prudential individual utility of a measure in such detail because it can be used to justify the most important premises in the argument for the actually central thesis of a justification of political measures. If political measures are about realising the common good and, as will be shown in a moment, the common good can be interpreted as the moral good, then in order to justify a measure it would have to be shown that the following central thesis applies to it:

Moral justification theses for political measures:

‘m1 is the morally best among the morally permissible and realisable (technically feasible and politically enforceable) measures (in the field of options under discussion).’

It is already clear from the meaning of this thesis that it follows deductively from a series of premises, i.e. it can also be justified in deductive arguments with these premises:

Premises for deductively justifying the justification thesis for a political measure:

M1: n premises of the kind: ‘The moral desirability of measure mi is vi.’;

M2: n premises of the kind: ‘mi is a realisable (technically feasible and politically passable and enforceable) measure in the field of options under discussion (it is e.g. a solution to a certain socially thematised problem).’;

M3: one premise of the type: ‘m1 to mn are all possible relevant and realisable measures in the discussed field of options.’;

(M4: n-1 premises of the type: ‘The moral desirability v1 of measure m1 is higher than the moral desirability vi of measure mi.' ('v1 > vi.’));

M5: one premise of the type: ‘m1 is morally permissible (i.e. does not violate other socially valid moral principles).’Footnote 4

The actually problematic among these premises, and the ones to be discussed further in the following, are those of type M1, i.e. the moral evaluations. The only theories that provide general criteria for the common good or for moral evaluations that can be applied to any political measures and that thus posit an overall societal (quantitative) target function for social action are welfare ethics.Footnote 5 In ethics, these are the standard approaches to the common good. Welfare-ethical moral evaluations aggregate the moral desirability of a measure from the various individual prudential utilities of this measure for all (sentient) beings, using certain mathematical aggregation formulas. The competing theories in normative ethics use different aggregation formulas here. The most important of these are the utilitarian, the prioritarian and the moderate egalitarian.

Utilitarianism says: the (utilitarian) moral desirability of a measure mi (VUmi) is equal to the sum of the various individual prudential utilities of mi for all single individuals (Smart 1973: Ch. 4):

Utilitarian moral desirability: VUmi := U1mi + U2mi + … + Uomi = j=1∑oUjmi,

with Ujmi being the prudential, individual utility of measure mi for the subject sj.

Utilitarianism does not take distributive justice into account. For utilitarianism, it is equally good if an additional unit of utility from an action benefits a person who is already doing very well, or if it benefits a person who is doing badly. Many find this unjust.

An alternative theory of moral desirability that attempts to specify this idea of distributive justice is prioritarianism, which says: "Benefiting [sic!] people matters more the worse off these people are" (Parfit 1997: 213). This means prioritarianism gives more weight to improvements for the badly off than to those for the better off, and the more weight the worse off the person is. However, the weight for improvements even for the worst off never becomes infinitely large—small improvements for them can be outweighed by many large improvements for the better off—; and the weight for improvements even for the best off never reduces to zero. Technically, this is realised as follows: In the prioritarian evaluation of a world situation (e.g. the world situation in which the measure m1 is realised: Wm1), first, the utilities of this world situation for the various individual subjectsFootnote 6 are each assigned a prioritarian desirability with the help of a (monotonously increasing) concave prioritarian evaluation function (PR(.), see Fig. 1). Because of the concave form of this evaluation function, equal increases in (prudential) utility produce less and less moral desirability increases, the higher the utility level of a person already is. Second, these moral desirabilities are added up. The prioritarian aggregation formula for the moral desirability of a world situation Wm1 is thus:

Fig. 1
figure 1

Prioritarian moral value function PR(u) (Lumer 2005b: 6)

Full size image

Prioritarian moral desirability of a world situation:

VP(Wm1) := PR(U1Wm1) + PR(U2Wm1) + … + PR(UoWm1).

Compared to utilitarian moral desirability, there are three technical differences: 1. Instead of directly using the individual, prudential utilities in the aggregation formula, these utilities are first evaluated with the prioritarian evaluation function PR in such a way that improvements for the worse off count morally more than improvements for the better off. To achieve this, PR is monotonously increasing and concave.Footnote 7

The x-axis of this curve represents the individual, prudential utility u; the y-axis represents the moral desirability of this benefit PRu. Due to the concave shape of this curve, increases in utility for the worse off, i.e. persons located on the left of the x-axis, generate greater increases in moral value than increases in prudential utility for the better off. 2. The second change is that, according to most prioritarians, the question of whether someone is well or badly off is not measured by how she is doing now or in this hour, on this day, etc., but rather by what the prudential utility of her whole life is. In moral evaluation, then, one must take into account the value of the whole life with the measure being evaluated, or, as is said here, the value of the world accompanying the measure. This world accompanying the measure m1 is designated in the formula as "Wm1". 3. Now, the value of measure m1 is not equal to the value of the whole world that goes along with it, but only the increase in the value of the world produced by m1. One must therefore subtract from the moral value of the world accompanying m1 the value of the world accompanying the zero action or zero measure m0, i.e. doing nothing. It is the same for all the alternatives to be compared. As aggregation formula for the prioritarian desirability of measure m1, one thus obtains:

Prioritarian moral desirability:

$$\begin{gathered} VPm_{1} : = PR\left( {U_{{1}} Wm_{{1}} } \right) \, + PR\left( {U_{{2}} Wm_{{1}} } \right) + \cdots + PR\left( {U_{o} Wm_{{1}} } \right) \hfill \\ \qquad\quad\quad -\, PR\left( {U_{{1}} Wm_{0} } \right) \, - PR\left( {U_{{2}} Wm_{0} } \right) \, - \cdots - PR\left( {U_{{\text{o}}} Wm_{0} } \right) \\ = \sum_{\mathrm{j} = 1}^{\mathrm{o}} PR\left( {U_{j} Wm_{{1}} } \right) \,-\,\sum_{j = 1} ^{o} PR\left( {U_{\mathrm{j}} Wm_{\mathrm{0}} } \right). \hfill \\ \end{gathered}$$

If a measure only changes the utility of n (n < o) subjects, then the moral desirabilities of the world with and without this measure m1 are equal in each case for all other subjects si (for all i > n: PR(UiWm1) = PR(UiWm0)). Thus, when calculating the moral value of a measure m1, according to the formula just provided, only the utilities of the n subjects really affected by m1 need to be taken into account.

Another justice-oriented alternative to the utilitarian aggregation of individual utilities is moderate utility egalitarianism. According to this criterion, not only should the sum of utilities be high, but the utilities of the different individuals should also be equal. The egalitarian value of a social distribution of utilities is therefore equated with the sum of all individual utilities, from which, however, a measure of inequality is subtracted. Such measures of unequal distribution are, for example, the Gini coefficient or the standard deviation, each multiplied by the sum of utilities.

How then to argue for a specific measure, for example a measure to contain some epidemic?Footnote 8 One would have to show that the respective measure is the morally best of the currently available and enforceable measures. For each option, the moral value would have to be determined by identifying what the major consequences are for individual groups of the population: health consequences through changes in the probability of becoming infected by the pathogen or developing certain symptoms; the costs of the actual medical measures; economic consequences through possible direct restrictions on economic activity or in the long run through increased taxes or greater government debt service or through economic disruption; social consequences through restrictions on otherwise normal activities, and so on. For all these consequences for different population groups, the individual utilities would have to be determined in terms of QALYs: How much and for how long do these consequences change well-being? The political enforceability and the legal and moral permissibility of the individual measures would have to be ensured. All this is very difficult to determine. But it is possible. And it would make explicit all those considerations that so far are suppressed in the usual discussion of such measures. This would require an interdisciplinary task force of perhaps 20 good specialists who would first have to develop a general empirical model with which the necessary data could be identified and from which consequences for the well-being of different population groups could be calculated; then new empirical premises would have to be entered into the model for each new alternative, i.e. set of combined measures, so that new alternatives could be compared with each other within maximally a few days. Politicians and the general public would have to be informed not only by means of a report with the complete calculation but also in short versions with explanations of the essential premises and results. This would give the critical public and politicians the opportunity to participate in the discussion about the premises and the normative specifications.

5.3 Criteria for Epistemically Good Arguments for Some Other Important Types of Theses in Public Argumentation

Arguments for other types of theses in public discussion cannot be discussed here in the same detail as the previous two; references to literature must suffice here (see already Sect. 3). However, comments on some of them are still needed.

T3: Arguments for empirical theses on natural, social, economic, individual, psychological consequences of political measures: According to rational hedonism as the theory of the prudentially intrinsic good (see Sect. 5.1 above) in conjunction with a welfare ethic, all morally relevant consequences must be pursued to their impact on individual well-being. At this point, therefore, very rich empirical information is required on how certain natural, economic, social, health, cultural, etc. situations affect the well-being of individuals. Respective studies are collected, classified and made accessible in the World Database of Happiness (https://worlddatabaseofhappiness.eur.nl/). Despite the approximately 16,000 collected studies on subjective well-being, there are two major problems with regard to the information needed for public arguments: 1. Most of these studies examine life satisfaction (‘how happy are you with your life as a whole?') and the like, but not affective, emotional well-being (e.g. ‘how good do you feel / how happy were you yesterday / just before answering this poll?'; 'which kinds of feelings of the following list did you experience yesterday: anxiety, anger, … joy, satisfaction …?'; or a more precise measurement: Kahneman et al. 2004a), as would, however, be necessary for the welfare-ethical evaluation. 2. Many of the analyses themselves, but above all the meta-analysis of the World Database of Happiness, concentrate on the correlation strengths between certain types of situations and well-being, but do not provide any information on the resulting (statistical) level of well-being or its change compared to the average of the total population, as needed for the welfare-ethical assessments. Here, therefore, is a huge research gap that needs to be publicised much more and requires a considerable reorientation of happiness research for its prudential and political-moral usability (see also: Kahneman et al. 2004b).

T6: Arguments for justifying the determination of the intrinsic good and for justifying moral principles: According to the above thesis (Sect. 2), the philosophical discussions about the content of the intrinsic good and about the correct criterion of the common good must also be conducted in the political public sphere. So far, this has virtually not been the case, among other things because of the abstract level of the discussion, but also because there is considerable disagreement in philosophy itself about what justifications, what types of arguments could even show what the correct determination of the prudentially intrinsic good and what the correct criteria for the common good or, more generally, what the correct moral principles are. For both issues, intuitionistic justifications are dominant at the moment, which, however, are unsatisfactory from an argumentation-theoretical point of view because they mostly beg the question. Within the epistemological argumentation theory, instead, instrumentalist approaches to justifying determinations of the prudentially intrinsic good (Lumer 2009: 241–388) and to justifying moral principles (Lumer 2009: 30–53; 2015) have been developed (overview of other justification approaches: Lumer 2009: 53–95).

T8: Arguments about the best strategy for politically pushing through a measure: The arguments discussed so far all concerned, directly or as sub-arguments for required premises, the selection of political measures to be implemented. A completely different topic of public arguments are strategies for politically pushing through already selected political measures: How can certain groups of voters be won over? How can a political party be persuaded to agree to a particular measure? These questions are not about whether a certain measure can be socially pushed through and enforced (see T4), but about what is the best strategy to push it through. Theses T4 are used as premises within an argument for a moral optimality judgement. This optimality judgement is about the moral improvement achieved by the measure, possible positive side effects and moral costs (through political concessions, consumption of resources of the political pressure group behind the proposal, etc.). The criteria for these arguments have already been discussed above (Sect. 5.2). In political discussions, there are many more types of arguments about best strategies. However, these concern best strategies from the point of view of a particular group. Such discussions usually do not take place in the general public.

6 The Factual Situation of Political Argumentation and a Strategy for its Improvement

De facto, the norms for epistemically good arguments outlined here are virtually not followed. For example, I have conducted an argumentation-theoretical analysis of the programmes of the major political parties for the 2017 federal election in Germany. All election programmes consisted for the most part of promises and announcements of what the respective party will do if it enters government. The proportion of argumentative text is low, less than 10%. The few arguments are almost exclusively simple end-means arguments: It is justified that the advocated measure is a sufficient or necessary means or a necessary precondition or a substantial part of the realisation of the desired goal state. This is far from the standards developed above. Arguments that approach these standards can perhaps be found at the level of parliamentary committees, expert reports commissioned by ministries; but even these arguments are only approximations. And the situation is probably no better in other Western countries.

Hence there is still a great deal to be improved in factual political argumentation from an epistemological point of view. In particular, the criteria for epistemically good arguments in politics also need to be much more widely known. The best contribution to this is probably that good examples, which initially come from science and applied ethics, are developed in the first place. The long and complicated arguments contained in them should be summarised in a generally understandable way. They should be made accessible to persons and institutions in the broad sense of the political public from whom a high level of argumentation can most likely be expected. If such examples are more and more disseminated, they could also encourage emulation. As a result of the improved discussion, the policies adopted could then also improve, as could the political and argumentation-theoretical education of citizens.