Which Democracy?

My starting point in discussing the notion of democracy is a widely shared view. It defines the aim (or the value) of a democracy as a process of giving the people of a particular country or region an effective say in deciding on all matters that do or may affect them in significant ways. The following five points briefly explain how this general definition can be further developed (for much more detail, see Radder 2019, chapter 6).

First of all, procedures of voting, which guarantee the people a say on which visions, policies, and implementations will be endorsed and employed in the governance of a country or region, constitute an essential feature of

Which Science, Which Democracy, Which Freedom? 117 any democracy. A common but stronger formulation (discussed in Kitcher 2006, 1212) is that democracy implies "having individual control” over the conditions that affect one’s life. This alternative formulation is based on the notion of individual autonomy. Although it may be quite common, it is less appropriate. The reason is that having an effective say in a collective process does not, or not always, imply being in individual control. I will come back to this point in the discussion of the notion of freedom in the next section. A premise of democratic governance is that most of the time most people possess, or are able to acquire, the basic capabilities required for democratic involvement.⁴ A somewhat weaker but more plausible claim is that working towards the realization of this premise is a crucial and worthwhile duty that deserves and requires an enduring commitment.

Deliberation constitutes a second important feature of a democracy. Because many visions, policies, and implementations are complex and have far-reaching consequences, high-quality deliberation requires knowledge and comprehension; and because the people involved cannot be supposed to agree on all relevant matters, a real democracy should provide concretely implemented opportunities for critical debate and citizen participation. This also presupposes the basic right of the freedom of speech (or, more generally, expression) and the support of independent public media. That is to say, a democracy cultivates institutionalized learning processes (including insights from a diversity of academic disciplines) to maintain or increase the quality of deliberation and policy making. These rights should not merely be proclaimed in an abstract way, but they must be guaranteed through a legally and institutionally supported public sphere. As Nancy Fraser (2014) argues, this public sphere should be actually, and not just formally, open to everyone; it should be critical, which requires examining the deeper roots (and not merely the surface phenomena) of relevant issues; finally, it should be effective, in the sense that its legitimate results have a real impact on actual policies, regulations, or laws. The significance of this second feature of democracy, deliberation, implies that elections during a state of emergency (in which these rights are strongly curtailed or even fully suspended) are invalid. A case in point are the Turkish elections for presidency and parliament held in June 2018 during an emergency state that had lasted for almost two years.

Third, the modern democratic state should be a constitutional democracy. Such a democracy is supported by a constitution and related laws concerning basic human rights and by the separation of the executive, legislative, and judiciary power of the state. An important normative aim of the constitution and the legal system is to protect the people as a whole against ill-considered policies resulting from the folly of the day and, in particular, to protect minorities against a potential tyranny of the majority.⁵

Related to the latter is a fourth important feature of democracy. It concerns the constitution of "the people”, the demos. In a traditional formulation, the core idea of democracy is “government of the people, by thepeople, for the people” (Mounk 2018, 56). The second and third points are crucial. Democracy means the rule of the entire demos. It requires that a democratic government cares for the interests of all the people affected by its politics, not merely for those of its own supporters.

A fifth and final point derives from the related question: who constitutes the electorate? Who has the right to vote? A principled answer is that the electorate should consist of all the adult, legal, and long-term residents who are living in a certain country or region. The reason is that they are the ones who are affected by the visions, policies, and implementations of the government. Thus, democracy requires universal suffrage of the electorate. Clearly, the electorate does not fully coincide with the demos, since the latter includes young children and minors, short-term residents, and people who lack a legal status. But although there are good reasons why these groups are not allowed to vote, they should be taken into account in democratic politics. First, because a democratic society has a duty of caring for their interests; second, because they may, and regularly do, make contributions to a flourishing democratic culture through their actions. See, for instance, the recent actions by school children and minors in raising awareness of the issues of climate change.

In sum, the value of democracy includes these five core features: voting, deliberation, a constitution, an inclusive demos, and an appropriate and inclusive electorate. As is generally true, values guide practices in certain directions (rather than others), but they do not determine the actual forms these practices may take. The latter requires additional interpretation and contextualization. Thus, actual voting procedures, forms of deliberation, legal systems, and definitions of the people and the electorate are not fully determined by the general value of democracy. For instance, although voting is an essential feature of democracies, its particular procedures should not be confused with the general notion of democracy as explained by its five core features. Particular voting procedures are pragmatic tools for decisionmaking processes. Put in philosophical terms (Radder 2006, chapters 8-11), the nonlocal meaning of the value of democracy is not identical with and exhausted by the actual voting procedures through which we attempt to realize it. Rejecting this identity is important, for three reasons. It urges us to acknowledge the question of how taken-for-granted procedures can be improved and made more democratic as a central challenge to democratic societies; furthermore, it encourages the exploration of different kinds of pragmatic democratic procedures (including well-thought-out and well-prepared referenda); finally, it underlines the crucial importance of creating and fostering a socio-cultural climate that promotes and encourages democratic engagement concerning all kinds of relevant public issues.⁶

The two following examples illustrate these points. The first is about voting procedures. It is fictive but cases like this one happen all the time. Suppose a government is based on the results of an election with a turn-out of 70% of the formal electorate. Suppose also that the government itself

Which Science, Which Democracy, Which Freedom? 119 consists of representatives of three political parties, one having ten votes in the Cabinet meetings, while each of the other two has five votes. Under these conditions, making compromises will be a frequent feature of the decisions made by this government. Thus, it will regularly happen that particular decisions are taken that are abided (but not really supported) by the two minority parties. In this case, these decisions are actually supported by 50% of the Cabinet members. Consequently, the decisions are based on no more that 35% of the full electorate. Put differently, in actual political practice many decisions are taken that are neither the will of the entire people, nor of the voters, nor even of the majority of the voters.⁷ Frequently occurring examples like this teach an important lesson, with theoretical and practical consequences. They imply that the notion of “the will of the people” is questionable and misleading. Theoretically, one might question whether it makes sense to assign some kind of “aggregated will” to collectives. More important is that in practical politics the phrase is frequently abused to legitimate decisions that are, arguably, not supported by the entire people. Therefore, in theoretical debates on democracy and in day-to-day practical politics the notion of the will of the people had better be left out completely.

The second example concerns the constitution of the electorate. The common view is that, during a long historical process, membership of the electorate has been stepwise expanded to include the elite and the masses, the rich and the poor, the male and the female, the white and the colored. In mature democracies, this process is taken to have reached its ideal final state of universal suffrage. “In a democracy ... all citizens get one vote without regard to the color of their skin or the station of their ancestors” (Mounk 2018, 130).

But is this really the case? The widely shared view of democracy, with which I started this section, implies that the electorate should consist of all the adult, legal, and long-term residents who are living in a certain country or region. In practice, however, in quite a few countries participation in (national) elections is not based on this kind of residency but on nationality or formal citizenship. This implies that, on the one hand, legal, long-term residents who (for some reason or other) do not have this nationality or citizenship are unjustly excluded from democratic rights. On the other hand, it inappropriately keeps granting these rights to emigrants who have been living abroad for many years, as long as they have kept the relevant nationality or citizenship. Depending on the specific countries, the two groups may have considerable sizes.⁸ These facts further detract from the representativeness of actual decision-making procedures, and so provide another reason to avoid the idea of the alleged will of the people. Instead, in a real democracy membership of the electorate needs to be based on residency, and not on nationality. In his recent book, Yasha Mounk rightly argues for the importance of an “inclusive patriotism” (2018, 207-215). A democracy requires the involvement of its citizens in the fortunes of thecountry or region in which they live. This involvement needs to be appropriate (thus, excluding those citizens who live abroad and are not, or hardly, affected by the relevant politics) and inclusive (thus, open to all the adult, legal, and long-term residents). Such an appropriate and inclusive suffrage would bring the realization of an inclusive patriotism one step closer.

Let us now turn to the subject of science and democracy. This is a complex subject that can be addressed from a variety of perspectives.⁹ In the second part of this section, 1 review a specific case of public participation in science policy and evaluate its democratic substance on the basis of the five core features of democracies explained in its first part. The case is the recent development of the Dutch National Research Agenda (DNRA). In Radder (2019, 244-251), I have provided a detailed account of this case, including the views of a variety of other commentators. Here, I first provide a summary and update of the implementation of the DNRA and then assess its democratic quality.

The aim of the DNRA was to consult the Dutch citizens about the question of which topics should be included in the research agenda of the Dutch scientists. It implied a stronger direction of the sciences toward central societal issues. The DNRA is claimed to be a unique form of citizen science, the results of which should show that science4competitiveness. The implementation of the DNRA, directed by a so-called Knowledge Coalition, started in 2015. The DNRA website summarizes the process as follows:

The Dutch approach to developing a national science agenda has been genuinely unique, in the way that the Knowledge Coalition invited the general public in April 2015 to submit questions about science. This resulted in 11,700 questions submitted by the general public, academic institutions, the business community and civil society organisations. Five academic juries were appointed by the Knowledge Coalition to cluster and assess the questions. This was followed by three conferences in June 2015 on “science4science”, “science4competitiveness” and “sci-ence4society”. Their purpose was to bring order to the questions, and to further aggregate the questions where possible, based on these three perspectives. This process ultimately led to the 140 overarching scientific questions and 16 example routes.

(Dutch National Research Agenda 2015)¹⁰

In agreement with the approach in this chapter, the notion of science is used in its broad European sense and the DNRA does not presuppose a strong contrast between science and technology.

To make the huge set of questions practically manageable required substantive processing procedures. Most of this incisive processing was done by the mentioned academic juries. It resulted in summarizing the 11,700 questions into 140 overarching questions. Based on the keywords added by the applicants to their questions, Maloti van Hintum (2015, 23-24) provides the

Which Science, Which Democracy, Which Freedom? 121 following rough but informative ranking of the topics of the most frequently posed 25 questions: health (480); energy (370); the brain (280); building (277); sustainability (270); education (260); the economy (242); development (212); behavior (192); chemistry (185); children (171); innovation (165); cancer (162); sustainable (161); security (156); prevention (154); language (147); climate change (146); data (146); sport (145); history (141); well-being (141); life (138); culture (135); society (128). In a first round of funding, the Dutch government made available a budget of 61 million euros. Teams consisting of scientists and commercial or societal partners have submitted research projects that address the overarching questions. These projects have been assessed by means of the usual science policy procedures. In June 2019, the results of this funding process have been made public: out of 323 proposals, 17 research projects will be actually funded.

Can this DNRA be seen as a successful form of citizen science (or citizen science policy) and, more broadly, as an exemplar of democratizing science? Unfortunately, this is not the case. The reason is that both its practical implementation and its conceptualization of the relation between science and society suffer from a variety of significant democratic deficiencies.

• 1 First, there is the composition of the Knowledge Coalition, and the Steering Group appointed by it, which is quite one-sided and biased toward economic interests. The Coalition consists of representatives of employers’ associations and of established policy organizations in the fields of higher education, technology, and medicine. Labor unions, NGOs, and grassroots collectives are conspicuously absent.
• 2 Second, the claim that “all stakeholders were involved” (De Graaf, Rinnooy Kan, and Molenaar 2017, 236) is not correct. At about the same time as the DNRA, a variety of academic grassroots organizations launched strong and broadly supported criticisms of the commodification, hier-archization, and bureaucratization of Dutch science and university policies.¹¹ These stakeholders have not been involved at all.
• 3 Third, there are two features of the practical implementation of the DNRA that significantly detract from its alleged bottom-up character. The first is that a substantial number of the questions (in some areas even half of them) have been submitted by professional scientists, either individually or as a group (Van Hintum 2015, 24-35). The second problem is that the process of clustering the 11,700 questions into the 140 overarching ones, which has been carried out exclusively by academics, was by no means a neutral procedure. The following example shows what gets lost in this process. One question reads: “How can we increase the inclusion, resilience and talents of young people with (mental, cognitive, physical) developmental problems and retardations?” According to one of the scientific jury members, the core of this question can be rephrased as: “How can we have children grow up safely?” (Van Hintum 2015, 51-52). But of course the original question is not limited to questions of safety. And while the latter question could be used to promote technocratic researchabout security policies, the former question could entail a rather different set of research projects.
• 4 The fourth problem concerns the way the issues have been split up in terms of the questionable distinction between science for science, science for society, and science for competitiveness, a distinction that is also used in current European science policy. Apparently, basic science is seen as “not for society”, social and economic issues are taken to be neatly separable, and the economy is reduced to competitiveness.
• 5 A fifth issue is the lack of reflexivity. The explanation of one of the questions states that the DNRA needs a more realistic view of what science can, and cannot, achieve. At the end of the day, however, this important request for reflexive understanding has hardly been honored: the final list of 140 overarching questions includes only two of such explicitly reflexive questions. One is about the normative issue of what is “good” science; the other about the possible limits of scientific knowledge (see Van Hintum 2015, 167-174).
• 6 A final problem concerns the funding procedures and the resulting decisions on approval and rejection of the submitted projects. The results of the first round of applications (which became available in June 2019; see NWO 2019) show the usual deficiencies and biases of this type of science funding (cf. Radder 2016, 93-101). For a start, the success rates have been extremely small: a mere 5.26% of the submitted projects have been approved. Furthermore, as can be gathered from the above-mentioned frequencies of the top 25 keywords, the full set of questions covers a remarkably broad spectrum of subjects that require both basic and application-oriented research from all academic disciplines. Beatrice de Graaf emphasizes the same point: “Among the ... queries submitted online, many asked questions having to do with the origins of mankind, with society’s resilience, with identity, democratic citizenship, and the need for spirituality and religiosity. Utilitarian motives did not predominate” (De Graaf 2017, 189).

In spite of this, the funded projects show a strong bias toward the technological, physical, and biomedical sciences. These disciplines received 90.5% of the available 61 million euros; left over for the social and human sciences was a poor 9.5% (Bol 2019). Moreover, “utilitarian motives” did predominate the funding phase: only a minority of the funded projects (my estimate is 5 or 6 out of 17) can be seen as primarily consisting of basic research.

What to conclude about the democratic quality of the DNRA? I will address this question with the help of the five core features of democracy discussed in the first part of this section. The main question concerning voting is this: did the general public have an effective say in the decisions concerning the content of the Dutch research agenda? The answer is: only in a very limited way. In principle, every citizen had the opportunity for submitting questions. But due to the specific institutional framing of the DNRA (points (1) and (3) above) its actual implementation included significant top-down

Which Science, Which Democracy, Which Freedom? 123 procedures and a clear bias toward economic issues.¹³ Furthermore, the summarizing and clustering by scientists has led to significant displacements and distortions of the original questions posed by the public (point (3)). Finally, although the decision-making processes about the funding of the submitted research projects have not been made public, such processes often include some kind of voting procedures by scientists and policy makers. Whatever the precise details of these procedures, their outcomes show a significant mismatch between the skewed disciplinary distribution of the approved projects and the much broader disciplinary variety of the topics of the questions submitted by the public (point (6)).

In the course of developing the DNRA, a variety of forms of deliberation have taken place, both in academic and in popular settings. Yet, the quality of these deliberations has been substantially impaired by the conceptual framing of the presentations and debates. As 1 concluded in points (4) and (5), the DNRA suffered from substantial conceptual shortcomings. They concern the implied divides between science, society, and the economy and the lack of substantial reflection on what we can reasonably expect from science for the purpose of solving our societal problems.

The third core feature of democracies is the support they receive from an independent constitution and legal system. In the case of the DNRA, this support is rather indirect. No special laws have been passed for the purpose of creating a Dutch research agenda. At most, one could say that policy making on an important societal issue like the social relevance of science is a legally supported task of a democratically chosen government. There is one specific question, though, that is relevant in this context. Although developing a DNRA has been broadly supported, there have also been criticisms claiming that it goes against the idea of academic freedom (Van Hintum 2015, 108-109 and 135—136; Molenaar 2017, 33—36). However, this criticism (which sees any public direction of academic research as illegitimate) can itself be questioned. First, as we will see in the next section, in the Dutch context the precise meaning and implications of the legal notion of academic freedom are unclear and disputed; second, the DNRA does not seem to impair a basic point of academic freedom, which is that it is still the scientists themselves who decide on the choice of the appropriate methods to study the relevant questions and on the acceptability of the proposed answers to these questions.

The two final core features are the inclusiveness of the demos and the appropriateness and inclusiveness of the electorate. In the first stage of the process, the collection of the questions, all citizens of the Netherlands, even children, were allowed to participate. However, this formal openness did not lead to a balanced inclusion of the entire demos (point (3)). Allowing the submission of questions by scientists, and even groups and organizations of scientists, has given this specific part of the public an unrepresentative and hence inappropriate weight. Furthermore, the summarizing and clustering was exclusively done by the members of the scientific juries, thus excludingwider participation of, and opportunities for genuine interaction with, the broader public at this stage. Finally, point (6) makes one quite suspicious about the representativeness of the (scientific) electorate of the committees that have made the final funding decisions. Although the detailed disciplinary composition of these committees and their external reviewers has not been made public, the hypothesis that it has not been proportional to the distribution of the questions over the different academic disciplines is quite plausible.

My conclusion from this analysis is twofold. Negatively, we have to conclude that calling the DNR.A a case of citizen science (or citizen science policy) is already questionable, but it is certainly not a successful example of democratizing science. Positively, this way of analyzing this attempt at democratizing science makes explicit the relevant issues that have to be taken into account and entails concrete ways of where and how possible future attempts should and could be improved.