Quest for Empirical Foundations of Deliberative Democracy Theory:
the Case of the European Union

Ph.D. Research Project for the EUI, Law Dept.

(This was my initial project, by now it developed in quite different and even slightly opposite direction)

I. Introduction: Theoretical Context

There are two major trends in the contemporary theory of democracy – the deliberative democracy theory and the social choice theory. The main tenet of the former is that through a process of rational public discourse a common will of the community is being formed and the legitimacy of the state and of the law is based on this common will formation. The most important characteristic of this deliberative process is that it is rational, and rationality is taken to mean generalizability in the Kantian sense, i.e. in order to qualify as rational any proposed rule should be generalised in such a way, that it can at the same time become universal. A proposal which cannot be recast in such a way can be dismissed as irrational. It follows that in the course of deliberation, the agents may change their preferences, values or judgements, yielding to the rationality of the proposals and the argumentation of the others, and this change is genuine. Thus, most versions of the deliberative democracy theory are normative rather than descriptive, claiming what the law, or the law-making process, should be in order to be legitimate, rather than explaining what they actually are.

The other trend – social choice theory, seems at variance with any of the above. First, it usually seeks for social choice function, which is an aggregation of a given set of individual preferences. While the process of deliberation is taken to produce something that is allegedly shared by everyone, the result of the mathematical aggregation may turn out to be different from the individual position of anyone. Second, it takes for granted a set of individual preferences, usually considered constant and unaffected by the process of decision-making. Finally, it is descriptive and claims to explain what does happen, and is rarely interested in what ought to happen. Yet in my view both theories have common ground in their shared reliance on rationality. Habermas believes that the rationality of the discourse will guarantee the eventual formation of the common will. On the other side of the divide Robert Aumann has proved mathematically that (in certain conditions) rational people cannot disagree.

II. The Proposed Research

The proposed research will seek to establish whether in several selected decision-making settings a common will of the participating agents is formed or the decision reached is result of mere aggregation of their preferences. It lies both in the field of jurisprudence and in the field of socio-legal studies. First, it will seek to redefine the deliberative democracy theory in such a way to make it falsifiable, then it shall turn to the empirical analysis of the European Union institutions, and finally it will evaluate the empirical results against the redefined theory to make further theoretical conclusions.

1. Identifying Observable Characteristics of Deliberative Democracy

The study will start with a survey of the deliberative democracy theory and will seek to identify several landmark features of deliberation, which are observable in an empirical research, and ‘measurable’ by the social choice theorists. I cannot be conclusive at the proposal stage, but I expect that these features will include preference change of the actors (the genuineness of the change may be ascertained by the consistency of their decisions in repetitive situations), consensual outcome even when rules allow voting (this is fairly common in the EU) and novelty of the outcome (this is important to distinguish whether new common opinion was formed or a coalition behind one of the initial positions was bargained). I assume that few such features will be sufficient to determine if the observed law-making process is an example of genuine deliberation or not.

2. Empirical Quest for Genuine Deliberation in the EU.

Having identified such features, the study will continue with an empirical research to test the deliberative democracy theory with a scrutiny of the process of legislative and administrative rule-making in the EU. This context is selected because of the observed “culture of inter-administrative partnership, which relies on persuasion, argument and discursive processes rather than on command, control and strategic interactions” (Joerges) and because in this case the deliberating agents are not individuals, but states and institutions, which presumably are more rational than individuals. Rule-making process at various levels and with the participation of agents with varying legal powers will be analysed, mainly from the perspective of the deliberative democracy, and the empirical observations will be evaluated against the several measurable characteristics of the deliberative democracy, identified in the first part. Social choice perspective will be applied also where it is possible to emphasise the effect of deliberation on the relevance of the conditions of certain social choice theorems.

This part of the research will consists of two sections.

(a) Analysis of the 'High Level' Decision-making: EU Legislative Process

The EU legislative process (the co-decision procedure) is a good candidate to have the characteristics of a ‘genuine deliberation’ because the Commission relies mainly on the rationality of its proposals and on the expert argumentation to challenge the entrenched national interests of the Member States, represented in the Council. The Commission has monopoly of initiative, which is a huge hard power and very important bargaining chip, nevertheless it has to persuade the other players, as well as the public, to adopt its proposals by means of rationalized arguments. In turn, the countries may defend their initial preferences only as long as they can rationalize them, i.e. recast them in terms of the general Community interest. For this reason the co-decision is an especially interesting test case for the ability of a deliberative process to result in formation of a common will and for the deterring effect the rationalizability requirement may have on the parochial interests backed by power. To the extend that the Commission is successful in getting its policy proposals voluntarily accepted by the initially hostile Member States (voluntary in the sense they have not vetoed the proposal, or have not built a blocking coalition against it) it may be proven that (1) deliberation at least in certain institutional contexts does create consensus and (2) does transform the initial preferences of the agents. This shall be evidenced by a detailed case study of the adoption of a controversial legislative proposal. 

(b) Analysis of the 'Low Level' Decision-making: Administrative Rule-making.

Another context that may present evidences for preference transformation is the administrative decision-making within the Commission or in the European Agencies – in cases of adoption of delegated legislation or legislative drafts or in cases of implementation in which the discretion is wide. Here the decision-making structure is inverted: a single player – say, a European Agency, has the sole power to set certain standard, while multiple other players – NGOs and lobby groups (whose ideal points may be at the opposite ends of a preference scale), are regularly consulted, but have power neither to block, nor to amend. The intuitive hypothesis here is that if the actors successfully rationalize their own position, i.e. if they present knowledge-based arguments to prove that the ideal point of the whole society is close to their own, the regulatory agency is likely to set the standard there, and more importantly, is likely to adjust its own position around that point. In cases of pure standard-setting, the latter is the only result that will support the deliberative democracy thesis, because there is no common will to be formed by the single regulatory authority, nor need the other actors be expected to change their own preferences. However when the agencies are drafting legislative proposals the issue is not one-dimensional and evidence for common will-formation may be expected and will be easily observable by tracing the origin of the rules that appear in the final bill.  

III. Study Outline and Research Questions

Account of the deliberative democracy and social choice theories (their versions, which are most suitable for the purposes of the research will be selected). From the deliberative democracy theory the study will identify several characteristics of genuine deliberative discourse, which can be used to evaluate the results of the empirical research. From the social choice theory the study will select few problems, which deliberative discourse may be expected to solve (possibly some of those pointed by Dryzek and List). 

Analysis of the decision-making in the EU on high and on low level:

(a) EU legislative process – one case-study of the interaction of the Commission, the Council and the Parliament on certain salient and controversial issue; and
(b) Administrative rule-making in the EU (within the Commission or in a European agency) – one case study.

The study of both levels is intended to answer the following narrow research question: are the observed processes instances of deliberative discourse or are they only methods of aggregation of preferences and bargaining.

Test of the empirical findings against the conditions and expectations of the deliberative democracy theory to determine the answer of the narrow question. Although the criteria for positive answer are yet to be determined (in part 1 of the research), this will probably be the case if there is observable preference transformation of the agents, consistency of their position taken in the case with their latter positions, consistency with their internal policies, the quality of the subsequent implementation of the decision by them, etc. Conclusions for the effect of certain procedural rules on the quality of deliberation may be made in passing. If the findings are positive, the case for empirical foundations of the deliberative democracy theory will be established.

Only if the conclusions of part 3 are positive, i.e. there are observed cases of genuine deliberation, an answer to a broader question will be attempted: how the observed evidence of an empirical deliberative process affects the selected problems of the social choice theory. This part will be theoretical, however it will be based on the observations of cases when collective will was indeed formed and actors did change their preferences. Thus, the analysis of the constitutional or administrative processes may prompt the need for the social choice theory to change some of its premises, i.e. to accept that (at least in some cases) preferences can genuinely be changed by the very process of decision-making or more generally, that the process of public decision-making is not merely a function that gives the product of initial constants, but is a social process, which itself affects the participating agents. The latter conclusion is in the same time trivial and radical. It is trivial, because it is intuitive that the process should affect participants and it is radical, because if proven empirically, it will require the social choice theory to change some of the main assumptions of its analysis.