Procedural Constraints and the Politics of Checks and Balances
Of course the lack of administrative clout, and even perhaps of fiscal discretion, would be of less consequence if the EU technocracy could act unhampered by procedural constraints. The EU's ability to act, even in those areas where it enjoys clear competence, is constrained by institutional checks and balances, notably the separation of powers, a multi-level structure of decision-making, and a plural executive. This makes arbitrary action--and, indeed, any action-- difficult and tends to empower veto groups that can capture a subset of national governments. Such institutional procedures are the conventional tool for protecting the interests of vital minorities-- a design feature generally thought to be most appropriate to polities, like the EU, designed to accommodate heterogeneous cultural and substantive interests. (Lijphart 1990)

The most fundamental constraint lies in the requirement of unanimity for amendment of the Treaty of Rome, followed by electoral, parliamentary, or administrative ratification--a high standard for any fundamental act of substantive redirection or institutional delegation. Accordingly, the EU has developed over the past two decades only by focusing on core areas of exceptionally broad consensus, backed by large financial side-payments to persuade recalcitrant member states. Whereas judicial decisions like the celebrated Cassis de Dijon case, which ushered in the single market, may have helped set the agenda for initiatives like the single market, monetary union or enlargement, there is now agreement in the scholarly literature that they could not do so without nearly consensual support from the member states (Alter 2002). Even "everyday" EU directives must be promulgated under rules that require the concurrent support of between 74 and 100 percent of the weighted votes of territorial representatives in the Council of Ministers--a level of support higher than required for legislation in any existing national polity or, indeed, to amend nearly any national constitution in the world today.

Surmounting supermajoritarian and unanimous voting requirements is not enough, however, to pass legislation. The EU is not a system of parliamentary sovereignty but one of separation of powers. Power is divided horizontally among the Commission, Council, Parliament, and Court, and  vertically among local, national, and transnational levels--requiring concurrent majorities for action. For legislation, the Commission must propose; the Parliament must consent; if the challenged, the Court must approve; national parliaments or officials must transpose into national law; and national bureaucracies must implement. Indeed, even within each branch and level of EU governance, we encounter extraordinary pluralism. The Commission itself is a plural executive--so much so that experts disagree whether it is an executive at all. The EP requires unusually high majorities to act. As a result, consistent and effective EU policy-making tends to be possible only where there exists not just a supermajority of national representatives but a supermajority of European technocrats, judges and parliamentarians as well. Current proposals to represent more groups, for example through another chamber representing national parliamentarians,can only exacerbate this tendency.

This is not the only passage in which the subject is discussed.

Your concluding comment appears to displace the discussion of "democratic deficit" to national state levels rather than the supranational level of EU governance. I would be just as comfortable to discuss Italy and Hungary on that matter vis-à-vis the Union as the recent discussion in a diary on the Hungarian situation. However, the discussion here is on a supposed "democratic deficit" on the EU level. My opinion as above is that I do not take the premise for granted nor do I find direct elections of a Union president of any utility or interest, all the more so to forge a "stronger Europe."