In 2009, the EU Committee of Regions adopted a White Paper on Multi-Level Governance. During the public consultation of this document, I prepared the following opinion:
From a theoretical perspective, the most convincing strategy of institutional design would be to ensure a congruence between those affected by policies and those eligible to elect the political representatives who decide on these policies. Such a congruence of constituencies would create the best conditions for policymakers to be held accountable for their policies and thus strengthen the incentives for responsive policymaking. In contrast, an incongruence between those responsible for and those affected by a policy would provide incentives for unaccountable and unresponsive policymaking (e.g. negative external effects, moral hazard, freeriding).
The congruence principle would imply that public functions are assigned to the territorial level (jurisdiction) that can be expected to (a) perform these functions most effectively, (b) comprises all important stakeholders and addressees of a given policy. Most importantly, congruence would suggest avoiding the sharing of responsibilities for policies between different territorial levels, because power sharing arrangements may dilute responsibilities and thus public accountability.
I see the subsidiarity principle as detached from its communitarian and organicist connotations and merely as an evaluative instrument (a heuristic) to achieve a clear assignment of functions to territorial levels, starting from the smallest jurisdiction. In principle, one could of course also start from the largest jurisdiction and ask whether smaller jurisdictions have comparative advantages in performing functions.
Thus, if multilevel governance means joint or nested governance, democratic theory would expect accountability problems to emerge. However, there are policy issues that can be most effectively addressed if different tiers of government cooperate. To ensure a maximum degree of accountability and participation, cooperation on multi-level problems should take place on a voluntary basis, not within an institutional framework that makes actions taken by one tier of government dependent on the approval of other tiers. In my view, the essence of the “partnership” model of governance suggested by the White Paper (p. 11) is (a) an equal democratic status of local, regional, national and European tiers of government and (b) the voluntary cooperation among these different tiers. Voluntary cooperation presupposes a mutual recognition of the partner’s democratic legitimacy. As democratic legitimacy does not necessarily require a “thick” identitarian attachment of citizens to one tier of government but is ensured through participation and accountability mechanisms, the EU tier of government relies on independent sources of democratic legitimacy and qualifies as a partner of equal status.
I do not assume that the current allocation of public functions to tiers of government in the European system of multilevel governance maximizes congruence of constituencies and would thus be the most appropriate institutional arrangement with respect to accountability or participatory democracy. To improve and revise the design, the principles of congruence and voluntariness (partnership) would suggest the following sequence of guidelines:
- All tiers of government should deliberate and decide whether a problem is best solved by one or several jurisdictions separately or whether it is a genuinely multi-level problem. Given the accountability and responsiveness advantages of separate-level policymaking, a separatory assignment of functions should be preferred if it is desired by one of the tiers. (This guideline would inter alia imply involving local and regional government in amendments of the Treaties.)
- If several tiers of government each claim exclusive responsibility for a policy problem, voluntariness suggests entitling each tier to formulate its own policies (concurrent powers). Varying legal rules may initially lead to conflicts, but these conflicts can be solved by courts or other dispute-resolution bodies representing the involved tiers. I would assume that learning and diffusion processes will, over time, lead to a convergence among tiers. The equal status and autonomy of each tier are likely to encourage cooperation between them.
- If cooperation fails and one tier of government disagrees with a policy implemented by another tier within the framework of concurrent powers, it may seek a revision of the other tier’s policy. However, the partnership principle would exclude forms of unilateral hierarchical instruction. Rather, the disagreeing tier should be required to obtain the support of a majority of the legitimating units within the other tier, eg. a majority within the European Parliament or a majority of regional assemblies, in order to induce the other tier to change its policy.
- If several tiers of government agree that the problem is of a genuinely multi-level nature and requires the cooperation between several tiers, a mechanism of power-sharing should be agreed. This arrangement should require all involved tiers to adopt policies jointly (for example, through contracts) and ensure joint accountability.
- If tiers fail to agree in the framework of a shared power, each tier should be allowed to opt out, reflecting the principle of voluntariness. In contrast with concurrent powers, the procedural threshold for a tier to deviate from a joint policy should be higher, as one can assume that decoupled policies are likely to produce less effective solutions for genuinely multilevel problems.
These guidelines are aimed at improving the institutional opportunity structures for democratic accountability in multi-level systems such as the EU. A crucial precondition for these guidelines to take effect would be the existence of horizontal associations of local and regional governments and, ideally, assemblies across the EU. The Lisbon Treaty fosters such horizontal association through its provisions on subsidiarity monitoring, and the White Paper suggestions are also very helpful in this regard. But existing horizontal associations (CoR, COSAC including second chamber committees, REGLEG, CALRE etc.) are still weakly institutionalized and lack capacity to participate actively in decisions on the assignment of powers.
Given the diversity of CoR members, which of course reflects the diversity of territorial state structures in the EU, I see the CoR not as a cohesive actor, but more as a network or interface for a variety of emerging horizontal associations among local and regional representative bodies. The “Governance” focus of the White Paper seems to acknowledge this role, but the White Paper does not explicitly reflect the role of the CoR in connection with COSAC, REGLEG and other associations of horizontal cooperation.
From the perspective of Germany as a country with powerful Länder capable of representing their interests through the nation state, multi-level governance is an attractive policy option only if the distinctive status of regions with legislative powers is recognized. The currently prevailing mode of Länder interest representation has been aptly summarized as “leave-us-alone”, i.e., urging the federal government to protect exclusive Länder powers against perceived EU incursions. The prevalence of this channel of interest representation also reflects the weakness of direct interest representation through the CoR. For a partnership-based governance to become attractive as an alternative mode of interest representation, the institutional representation of Länder and other legislative regions needs to be strengthened. This does not preclude strengthening the CoR or the representation of municipalities / municipal associations from unitary states, but it suggests thinking about a more differentiated structure as a precondition for partnership.