“Daring to Be a Daniel”

A Salutary Note of Caution, But

There is no doubt at all that Matthew Flinders’s (2011) article is an incisive and thought-provoking piece on possible pathologies of democratic accountability. Probably, only extremely populist-minded people would dispute the normative foundations of its (severe) diagnosis. Here, by “populist” I simply think (without any negative connotation) of people who—in a Rousseauist vein ¹ —profoundly distrust the fundamentals of representative democracy and most notably the “perils” (Lupia, 2003) lying in the delegation of popular sovereignty to professional politicians. In such a view, the latter are strongly suspected of confiscating sovereignty to benefit their own interests (or the interests of narrow constituencies on whose support they depend) or, to use terms of “principal-agent” theory, they are expected to “shirk” if their behavior is not monitored efficiently. For those (like me) not belonging to this category, it is not difficult to share the article’s skeptical conclusions about the idealization of accountability that becomes thus “the über-concept of the late 20th and early 21st centuries” (Flinders, 2011, p. 597, italics in original) or at least, in Pollitt and Hupe’s words, a “magic” concept among others (Pollitt & Hupe, 2011, pp. 647-649).

The plea for “proportionality” in the use of accountability is wise: Governance is “the art of gentle combination and alternation of trust and control” (Edelenbos & Eshuis, 2011, p. 23). As the author acknowledges, such a view is not new. When Flinders suggests that excess accountability can be counterproductive, he follows several other prominent theorists of accountability. Hood (2007, pp. 202-203) summarizes the perverse effects of accountability under Albert Hirschman’s concepts of futility, jeopardy, and perversity. Bovens (2005) emphasizes the risks of “excess accountability,” and similarly, Heald (2006) signals “the danger of over-exposure” (p. 60). Few would disagree that the necessity to continuously report on one’s deeds and justify them under the threat of sanctions may lead to a feeling of suffocation that induces subterfuges and blame-avoidance strategies. Risk-adverse individuals acting rationally normally want to shield themselves from accountability and will develop attitudes of excessive proceduralism, which can paralyze creative thinking. Further support could be provided to Professor Flinders’s claims by the work of other distinguished scholars suspicious of accountability. For instance, Philp (2009) alerts us to the risk that if accountability demands are too pressing, only ritualistic “appearance of conformity” to the expected behavior will be generated (p. 43). Examples thereof abound in empirical works of organizational sociology, which regularly emphasize spiral effects of control leading to unintended behavior, which nurtures demands for further control, and so on. To cite yet another body of research, studies on transparency and publicity (which are preconditions for accountability) conclude that their effects are ambivalent (Naurin, 2007). For instance, publicity makes it more difficult for actors to invoke justifications for their options and policy choices that are based on self-interest—this has been eloquently portrayed by Jon Elster (1998) as “the civilizing force of hypocrisy” (p. 12)—but publicity can also have deleterious effects on the quality of deliberation, for example, by favoring “plebiscitory reason” (Chambers, 2004) or by preventing the forging of compromises that are necessary for problem solving (Spörer-Wagner & Marcinkowski, 2011).

Notwithstanding these undeniable merits of Professor Flinders’s critical piece on the pathologies of accountability, I am afraid that there are some weaknesses in his line of thinking. These weaknesses regard the etiology of the disease, the symptoms of the pathology, and its consequences for the overall health of democracy. They are concentrated, in my view, in two partly inaccurate descriptions: one on the role of the media and one on the impact of “monitoring” in contemporary governance.

The Negative Impact of Mediatization: Exaggerated?

According to the author’s diagnosis, one major cause of the problem lies in the destabilizing role of the media. This argument is not entirely new either. Long before “mediatization” of society became a core topic in social science research, the report to the Trilateral commission by Crozier, Huntington, and Watanuki (1975) strongly criticized the negative impact of the media (together with that of intellectuals) on governability. ² Be that as it may, more recent empirical research in the field of communication science confirms Flinders’s negative judgment that “mediatization” (Esser & Pfetsch, 2004; Schulz, 2004) combined with media commercialization induces cynicism or distrust of politicians. Longitudinal studies suggest that in different countries, the content of the news about the political process has become more negative and more critical of politicians (Dalton, 2004, p. 71). Coleman and Blumler (2009) write,

The present system disseminates an oversupply of oxygen for cynicism through the visibility of manipulative publicity efforts and the increased flow of negative messages. A related product is a highly pejorative, oversimplified and, in many cases, probably unfair stereotype of the standard politician as someone who cares only for power and personal advancement, is not bothered about problems that matter to ordinary people and is constitutionally incapable of talking straight. Audience disillusionment with political leaders and their utterances is a natural outcome of such a bombardment. Credit is not given to government even when it deserves it. (p. 60)

However, comparative works also qualify Flinders’s argument, which cannot be generalized. Many cross-national differences persist in the architecture and practices of media systems and in the communication cultures both of journalists and politicians. Hallin and Mancini (2004) distinguish between three types of media systems: the liberal model, which prevails in Britain, Ireland, and North America; the democratic corporatist model, which prevails across northern continental Europe; and the polarized pluralist model, which prevails in southern Europe. Briefly stated,

The Liberal Model is characterized by a relative dominance of market mechanisms and of commercial media, the Democratic Corporatist Model by a historical coexistence of commercial media and media tied to organized social and political groups, and by a relatively active but legally limited role of the state; and the Polarized Pluralist Model by integration of the media into party politics, weaker historical development of commercial media, and a strong role of the state. (Hallin & Mancini, 2004, p. 11)

The degree of commercialization of the media system (“Americanization”) seems to be the outcome of “path-dependent” mechanisms—that is, to depend on long-term national traditions and trajectories. Therefore, convergence toward “Americanized” media practice—conducive to sensationalism, conflict dramatization, a culture of negativity, and attraction to scandals and policy failures—remains limited in several media systems outside the Anglo-Saxon world. Rather than homogenization, one may be tempted to speak of the “hybridization” of media regimes: a combination of characteristics influenced by the American model with country-specific traditions (Esser & Pfetsch, 2004, p. 406). The specificities of media regimes matter in turn for the political consequences of mediatization: In continental Europe—and especially in Scandinavian countries—these regimes are more favorable to a “public service” model that privileges hard news over commercialization so that the devastating consequences for the political culture due to the role of the media that Flinders observes in the United Kingdom are not necessarily observable (at least to the same degree) elsewhere as well. In sum, my claim is that the diagnosis of a pathology due to the media role cannot easily be established about democracy in general. It is based on the incorrect assumption that media systems without exception operate according to the same logic as the Anglo-American (“liberal”) model that the author has in mind (and from which he draws his evidence).

But even if a media system operates on these lines, as is the case in the United Kingdom (there is no reason to dispute the author’s evidence and arguments on that point), does it really produce the alleged negative effects? Professor Flinders posits with vigor a straightforward negative causal relationship between the increasing role of accountability hanging as a Damoclean sword over the shoulders of politicians—in relation to the role of the media as accountability “forums” (Bovens, 2005)—and the “evisceration” of public trust. This is even the core thrust of the argument in the article. Nevertheless, the empirical evidence that Flinders provides in support of his claim is unconvincing. The author refers to the whole range of comparative evidence based on survey research on the decline of public trust. However, this kind of evidence, although compelling in the description of the dependent variable (the amplitude of distrust), does not support Flinders’s conclusions about the explanatory factor(s) because it hardly suggests anything about them. Causal mechanisms are more apparent in Flinders’s interpretation of the decline of public trust in politics in the United Kingdom following the scandal of parliamentary expenses (Flinders, 2011, pp. 603-604). However, even if we charitably consider this “extreme example” (as acknowledged by Flinders) to be more generally relevant, correlation does not necessarily mean causality. If one tries to reconstruct the causal chain of events having led from media reporting of scandals to the erosion of trust in politicians, one realizes that there are some elements missing from this chain, notwithstanding the admittedly broad popularity of the “media malaise” and “spiral of cynicism” theses. I have no difficulty believing the data about trust erosion in the United Kingdom or that the observed abrupt decline of trust is largely due to the scandal. My reservations are of another nature.

First, we know that public opinion is volatile, and there is no evidence provided that the decline in trust is durable, that is, the scandal as a critical event (dramatized by the media) had a structural path-shifting effect on public confidence. To substantiate this, we need more temporal distance. More fundamentally, regarding causality, the attribution of responsibility to the media is delicate. ³ The author implicitly assumes that people are subject to the strong manipulative influence of negative reporting by the media—or, to state it a bit more bluntly, that they are “media victims.” In my view, this belongs to the “self-evident truths” of which Flinders is (correctly) suspicious. Actually, the relevant literature also provides indications to the contrary. Citizens are now globally better educated and have a greater interest in politics (“cognitive mobilization” thesis), which may predispose them to not be “media victims.” Furthermore, to believe the media, people need to trust them, but it appears that the level of trust the media enjoys is pretty low (Newton, 2006, pp. 216-217). In the United Kingdom, where tabloids dominate the market, the proportion of those tending not to trust the press reaches 79% (Eurobarometer 2010). Of course, one could argue that in spite of the overall trend of cognitive mobilization, the least educated segments of the population are more prone to becoming “media victims,” and thus, “attack journalism” undermines trust in politicians at least within these segments. However, these segments tend to lack trust in politics to begin with. For example, a study of the American presidential election of 2000 showed that those with low levels of political trust do not become more or less trusting following news exposure, regardless of the news source (Avery, 2009).

I do not contend that these counterclaims are more convincing than those made in “Daring to be a Daniel.” I am just suggesting that there is no sufficiently robust empirical evidence so far proving that media effects on individuals are stronger than in the past or that they are ubiquitous across national democratic systems (see Newton, 2006). Hence, we cannot demonstrate the negative impact of the role of the media as “watchdogs” for the legitimacy of democracy in general. Accountability mechanisms are control devices, and the interplay between control and trust is complex. There is no theoretical agreement on the nature of this mutual relationship, as suggested by another article recently published in this journal (Edelenbos & Eshuis, 2011). Arguments on the negative impact of excess accountability should take stock of theoretical developments on this topic and be considered as hypotheses that require empirical testing through “process-tracing” methods (George & Bennett, 2005) aiming to rigorously establish causal relations. For example, Flinders (2011, p. 606) suggests that the information processed in accountability processes “becomes politicized and therefore amplified which, in turn, further erodes public confidence in politics and fuels demands for the creation of even more elaborate accountability networks.” In such a narrative, there is a linear sequence of four causal effects of accountability: (a) politicization, (b) amplification, (c) erosion of public confidence, and (d) new demands for accountability mechanisms. One needs indicators not just for public distrust (Sequence c) but also for the other sequences, as well as some compelling evidence about their causal relations. Further research is thus needed to confirm the existence of the causal chain. I would frame it as follows: “Daring to be a Daniel” offers much food for thought that is extremely helpful for the formulation of insightful inductive hypotheses, but generalizations without more substantial empirical testing of such hypotheses are not very prudent.

The Penetration of Monitory Demo(?)cracy: Exaggerated as Well?

Another source of democratic malaise lies, according to Professor Flinders, in the advent of “monitory democracy” (Keane, 2009), in which the action of political officials is under the close scrutiny of different sorts of groups and institutions. It is no accident that in his seminal book on the transformations of representative government, Bernard Manin (1997) describes the current period as that of the advent of “audience” democracy.

However, the impact of public scrutiny on decision making is overstated in Flinders’s piece. Monitory democracy faces structural limitations and therefore is not as obtrusive as suggested. Obviously, mediatization is instrumental for monitoring, but it does not entirely permeate the political process. The media logic colonizes the part of political activity that is more prone to being mediatized: “frontstage” politics. This most visible part of politics can more easily become the object of dramaturgy, but this is not all the essence of politics. Policy making usually appears as a complex process entailing negotiation and compromise, and the involvement of the bureaucracy as well as of organized interests and other nonelected actors. True, politicians are under increased pressure to take into account the short-term media agenda: They need “something to show now” (Pollitt & Bouckaert, 2004, p. 8). Yet, policy making also implies long-term considerations, as it requires some continuity, and there are influential actors, such as those in the bureaucracy, for whom media constraints are not the primary concern. Moreover, the degree of complexity of policy-making activities is probably increasing. Hence, there is a broadening gap between mediatized “frontstage”—mostly partisan—politics and the more complex policy-making activities, a significant part of which mostly takes place “backstage.” Although “it sometimes seems that ‘politics’ is what appears on TV or in the press” (Street, 2001, p. 5, italics in original), this is not a general rule. As suggested by Hajer (2009):

While media attention gives a boost to the authority of the centre, actual problem-solving requires complex forms of network governance…. Hence, while the autonomous power of the centre to change events diminishes, its press coverage soars…. Network governance features those who know most about the substance, not those who like to be on stage. Should some enlightened spirit try to connect that roundtable to the mediatized politics, all that network governance can offer is a meeting room filled with administrators, some stakeholders, expert reports, and a folder with minutes and draft agreements. This can hardly be made into an interesting story, and is therefore neglected by journalists and remains invisible to the general public. (pp. 176-177)

True, it may be argued that scholars like Hajer overestimate the presence of “network” forms of governance and that other bodies of work tend to rehabilitate the “authority of the centre” (such as research on the “presidentialization” of politics; Poguntke & Webb, 2005). Nevertheless, a whole range of governance activities escapes public scrutiny, and significant aspects of policy making escape media attention simply because they are not interesting in commercial terms or because journalists lack the background to apprehend them. This is all the more the case with the internationalization of rulemaking.

The accountability of officials participating in international negotiations is to a large extent fictitious because those (like national legislatures) who have the formal power to hold them accountable seldom possess the necessary information to do so. International rules are prepared by top-rank administrators and officially negotiated by members of the executive, with government officials mutually engaged in cooperative relations of network type (Slaughter, 2004). These relations remain a black box for outsiders, and the relevant literature has signaled the “two-level games” (Putnam, 1988) that informational asymmetry allows the negotiators to play. In this case, one could hardly speak of excess accountability, but rather of significant accountability deficits. Such deficits are accentuated by the fact that most international organizations are hybrids, incorporating a negotiation system comprising representatives from not only national governments but also a less visible global body (for instance, a secretariat) acting to a large extent autonomously (Mayntz, 2008, p. 52).

Cooperative governance at the international level also takes place in less formalized, specialized instances such as the Basel Committee on Banking Supervision (BCBS), the International Organization of Securities Commissions (IOSCO), the International Association of Insurance Supervisors (IAIS), and the International Competition Network (ICN). These organizations “tend to operate with a minimum of physical and legal infrastructure; most lack a foundational treaty, and operate only along a few agreed upon objectives or bylaws” (Slaughter, 2004, p. 48). Although such instances do not have the capacity to issue binding decisions, it is hardly questionable that their national members—who do have such a capacity—are strongly influenced by their debates. Considering their influence in cross-national policy coordination and convergence (“soft law”), the fact that such networks are composed of transnational elites raises questions about their accountability similar to those raised about formal international regimes. But the fact that informal networks operate in a sort of “gray zone” and are not necessarily composed of formally authorized representatives is a supplementary source of concern.

In addition, the global level appears to be the level with most “islands” (Stone Sweet, 2004) of private governance giving a key policy role to democratically unaccountable instances. Private international regulatory regimes exist today in fields as diverse as regulation of the Internet and intellectual property, the international minerals, insurance, maritime transport industries, and industrial production standard setting (Hall & Biersteker, 2002, p. 30). One can thus speak of “the construction of many public policies as a result of what are essentially private efforts” (Hirschland, 2006, p. 12). To be sure, each policy sector regulated at this level has its own distinct decisional architecture, in which the roles of public and private actors, the divisions of competencies across territorial levels, and the intensity of regulation vary (Mayntz, 2008, p. 53). Nevertheless, there seems to be an important commonality: probably even less than in other forms of governance beyond the nation-state, private governance activities are not much the focus of monitory democracy. By ceding to methodological nationalism, we run the risk that national “frontstage” obscures the transnational “backstage.”

A last point that deserves attention is that with monitory democracy we might face the paradox of perhaps more accountability (however, not beyond national borders) coupled with less democracy. In a democracy, the monitoring agents are the citizens who compose the demos. But in “monitory” democracy, this is not true any longer. This is primarily “advocacy” democracy (Dalton, Cain, & Scarrow, 2003. p. 267 ff), in which the monitoring agents (e.g., courts, auditing institutions, etc.) are alerted by all sorts of “whistleblowers” (e.g., cause groups, the media, etc.) who claim to represent the people or to act in their interests. There are basically two problems with respect to the democratic credentials of such an accountability regime. First, monitoring requires resources that are unevenly distributed, so the (at least formally) egalitarian democratic principle of “one man, one vote” vanishes. In the nonelectoral domain, there is no equality of voice and “the advantages of education, income, and other unequally distributed resources are more likely to translate into patterns of over- and underrepresentation” (Urbinati & Warren, 2008, p. 405). To state it a bit differently, even in democracy without adjectives, “votes count, but resources decide” (Rokkan, 1966, p. 105), but in “monitory” democracy, votes count less and resources decide more. Second, self-proclaimed representatives (such as NGOs) that press decision makers to render accounts act as “surrogates” (Rubenstein 2007, p. 625) for the populations whose well-being is of concern to them and as instances of surveillance in the name of these populations. Yet, the latter have not delegated this task to them, nor are they able to sanction them if they are not satisfied with the way they perform it (Sperling, 2009, chap. 6). It may even happen that the “represented” are not informed at all about the actions (like monitoring of decision makers) undertaken “in their favor” or may ignore the sheer existence of such alleged defenders of their cause. Needless to say, the media contribute to the “labelization” of self-imposed representatives, especially when established channels of representation are lacking (Hajer, 2009, p. 183). Flinders (2011) sees well the instrumentalization and politicization of accountability by sectional interests and unaccountable elites in monitory democracy (pp. 608-609), but I think one could be more assertive on the undemocratic essence of this form of monitoring.

In his article, Matthew Flinders views monitory democracy as a dangerous evolution that should be feared (in contrast to John Keane, who celebrates its advent). On the basis of the above developments, I would rather point out that there are a number of (“backstage”) governance areas in which more “monitoring” would not cause harm to democracy but would rather be beneficial for democratic control. However, this monitoring would not be satisfying in terms of democratic quality unless it became more pluralistic and representative and its promoters became more accountable themselves. In my view, the major current pathology of accountability is not that there is too much monitoring in policy making but that there is an increasing number of governance areas in which there is either too little monitoring or monitoring that lacks democratic credentials.

“Fine-Grained” or “Broad-Brush” Analysis?

Professor Flinders claims that his approach contributes to a more “fine-grained” diagnosis of the (negative) evolution of democracy. Probably (I refer here to self-assessments by the author) one cannot be “provocative,” use “a fairly broad brush,” and produce “fine-grained” analyses altogether. I rather tend to believe that the author’s diagnosis overstates some of the (negative) phenomena that are observed. The assumptions about the etiology of democratic pathologies are not always robust, the symptoms are dramatized, and consequences are treated as if one would be in presence of a sort of large “pandemy” (to continue with the medical metaphor), notwithstanding the limits of legitimate generalization.

I am not sure that I would be wrong if I said that “Daring to be a Daniel” is an expression of anger and emotion because politicians are unduly “mobbed.” I feel much sympathy for that view, and we have all heard of cases of unjust verbal lynching and harassment by the media. Now, Flinders seems to adhere to the idea that “most politicians and public servants are not corrupt but enter public service out of a sense of civic duty and belief in collective endeavor” (Flinders, 2011, p. 612). Again, this is a (probably overly optimistic) generalization about something that is very much subject to empirical variation. Although I agree with Flinders’s positive evaluation of politicians in the United Kingdom, I have come to realize in my everyday life, not as a political scientist but as a Swiss and Greek binational “lay” citizen (two countries with very different corruption levels), how much the quality of politicians’ ethos (and thus the necessity to replace trust with control) is a variable even within democracies. Anger and emotion are legitimate feelings, and they can even be fruitful for hypothesis generation. However, they are not of sound advice when they provoke inferential jumps from single cases or episodes to general diagnoses of the health of democracy.

Finally, I also have much sympathy for the article’s conclusive statement about the responsibilities of the political science and public administration communities. The increased closure and specialization of these disciplines is prejudicial to the public engagement of their members. Scholars are too often obsessed with a narrowly defined scientific excellence to the detriment not only of the form of their message but also, more fundamentally, to the detriment of the social relevance of their research. Professor Flinders is a genuine, publicly minded intellectual, well aware of the normative requirements of his social role, and who should be sincerely applauded for privileging in his work the “logic of appropriateness” over the “logic of consequentiality” (March & Olsen, 1989). If others would emulate him, this would no doubt be an improvement because more public involvement of political scientists is a necessary condition for an evidence-based assessment of the quality of our democracies. The question is whether it is a sufficient condition. In my experience, the problem is not only that involvement in public debates does not feature as a high priority for utility-maximizing scholars primarily interested in making an academic career but also that decision makers, “stakeholders,” and the public are seldom keen to be “enlightened” by findings from scientific research. If a socially receptive context is required for accusations of corruption to find resonance (Flinders, 2011, p. 605), the public is less receptive to the more dispassionate discourse of science. Unfortunately, the resonance of our message does not only depend on its properties (that we can shape) but also on the dispositions of the target audience as well. Similar to the media, we are dependent on our public.