The EU’s Democratic Deficit is Structural not Reformable

Prof. Richard Rose |

The European Union’s founders learned about, long before democracy became the touchstone for legitimacy. In the words of Jean Monnet, the behind-the-scenes organiser of the treaties, ‘I thought it wrong to consult the peoples of Europe’ since they had no practical experience of diplomatic negotiations and politics (1978: 367).  European institutions agreed between national governments were legitimated on instrumental grounds: preventing another European war and promoting economic growth. The treaties gave legal legitimacy to the many small-scale policies promoted by technocrats

The legitimacy of government in the twenty-first century depends on it being democratically elected. National governments are democratic because they are accountable to parliaments elected by universal suffrage. Citizens thus have multiple means of showing their dislike about what is done in their name. They can vote to remove from office a national government that has endorsed an EU policy that is nationally unpopular and they can elect Eurosceptic MEPs. Citizens can also vote for referendum measures that challenge the EU’s authority.

It is impossible for the European Union to have the same democratic legitimacy as that of member states, because the obstacles to a democratic EU are structural. In the election of the European Parliament, the EU cannot apply the basic democratic principle of one person one vote, one value, because of gross inequalities between the population of member states. In any case, the EU is not a creation of the people of Europe. It is a league of member states, all of which are equal in law regardless of their population. The treaties that serve as its constitution are created by deliberation and bargaining between leaders of member states. Policies that the multi-national Council of Ministers and Parliament EU policies cannot be altered by the outcome of an election. The EU doctrine of the acquis communitaire means that once a policy or treaty is adopted it is not subject to repeal, unlike a national constitution or national Act of Parliament. The European Union’s political system is best described as ‘democracy lite’ (Rose 2015: chap. 4). Given structural obstacles to full democratic representation of its citizens, the European Union is not a political system that is democratising; it is a system with a built-in democratic deficit.



A state requires more than free elections to be fully democratic. Its elected government must also respect the rule of law and be subject to checks and balances between its political institutions. EU treaties grant equal legal rights to every member state and give smaller states disproportional voting strength in policy-making. Smaller EU member states are collectively insistent on enforcing procedures and laws that favour them against abuse by the largest member states. EU permanent officials prefer to be called bureaucratic because strict adherence to EU laws is their best protection from undue influence by national governments wanting to bend laws to their national advantage. In cases of major dispute, the Court of Justice of the European Union has the power to make and enforce decisions to prevent the violation of EU laws.

Within the European Union system, there are a substantial number of horizontal checks and balances. The adoption of policies requires the approval of both the Council of Ministers of national governments and the European Parliament. Within each institution, there are checks to prevent dominance by one or a few populous countries. The political heads and supranational civil servants of the European Commission propose and administer EU laws, but their actions are subject to monitoring by other EU institutions. The checks are so numerous that ambitious EU officials claim that they delay or prevent necessary decisions being made.

To be fully democratic a political system needs the vertical check provided by elections held on the basis of one person, one vote, one value. The EU cannot hold such elections because it is a union of states extremely unequal in population. Germany has more than 150 times the population of Malta and almost ten times the population of states with the median population, Portugal and Sweden. Smaller states are the norm; more than two-thirds have national populations below the EU arithmetic average of 16.5 million. Smaller states are very unequal in population too. They range from Belgium, the ‘biggest’ of the below-average states, to eight countries each of which has 1 per cent or less of the EU’s population. Four states – Germany, France, Italy and Spain – collectively have a majority of the EU’s population. To protect the majority of member states from domination by a few countries, EU treaties confer many rights of representation equally on each member state regardless of its population.

In the European Council nationally elected heads of governments collectively discuss EU policies of highest importance. Each state has one seat which is taken by its prime minister. In parallel, the Council of Ministers deals with issues of political concern to a specific government department. Gross inequalities in population between member states mean that the Council does not take decisions by a majority vote of states nor are votes weighted by population. Instead, endorsement of a policy depends on a super-majority weighted by both the number of states and by population. In a 27-member EU, a minimum of 15 states must give their approval. This means that two-thirds of the votes cast to approve a policy must come from states ranging in population from average to tiny. In addition, approval requires endorsement by states that collectively have at least 65 per cent of the EU’s population. Thus, a policy can be adopted only if it is supported by the biggest states and by many small states. Unanimous approval is required for a small number of major issues. The unequal representation of citizens and the requirement of unanimity are inconsistent with the democratic practice within member states.

The significance of inequalities in Council votes is mitigated by the great majority of decisions being taken by a consensus arrived at without a vote. Before putting a proposal to the Council for approval, Commission staff sound out country representatives most likely to be affected in order to modify features that they would object to. Upwards of four-fifths of Commission proposals have so little political impact that they are agreed by committees of national civil servants without reference to their national ministers. When Council members discuss major policies, participants are well aware of the scale of resources that big states have and of the seriousness of a particular issue to some small states. The upshot is that a bargain usually gives almost all states something but not everything that they want. If there is no consensus, then Council members can recommend that a proposal be withdrawn to avoid a bitter conflict (Thomson 2012).

The European Commission is the approximate equivalent of the executive branch of a national government. Each directorate is staffed by a multi-national civil service committed to European integration and recruited through rigorous competitive examinations (see. Kassim et al. 2013). Four-fifths of the EU’s laws and regulations are technical and limited in impact; they are approved by discussions between the Commission and national civil servants. Cumulatively, the spillover of lots of little regulations advances European integration (Niemann and Schmitter 2009). Since 1979 members of the European Parliament (MEPs) have been elected in 27 national constituencies. For MEPs to represent the views of voters, a significant majority of the electorate ought to vote. At the first EP election the nine member states saw a turnout of 62 per cent. In 2014 it hit its lowest level, 42.6 per cent, before rising to 50.6 per cent at the 2019 election, a figure still well below national levels. Until challenges to EU policy have been raised in the past decade at national level, low turnout has been interpreted as indicating passive and permissive consensus to EU policies (Lindberg and Sheingold 1970: 41).

Although the European Parliament is popularly elected, the formula for allocating MEPs to countries is a structural obstacle to the equal representation of European citizens. Seats are allocated by a system of degressive proportionality that turns proportional representation upside down (Rose 2015: 102ff.). The smaller the population of a member state, the fewer the number of electors required to obtain a seat; moreover, the least populous states are each guaranteed a minimum of six MEPs. The result is disproportional representation. In the 2019 European Parliament election, Malta had one MEPs per 70,000 people while Germany had one MEP for 860,000 people. Twenty-two member states had more MEPs than would have been the case if seats were allocated by the proportional representation of European citizens.

Within each national constituency, the EU requirement of proportional representation fragments the representation of national interests among up to half a dozen parties. The combination of proportional representation within national constituencies and disproportional representation between countries results in upwards of 200 different national parties being represented in the European Parliament. However, only seven parties have won as much as 33 per cent of their national vote, and only four parties have returned at least two dozen members. The two largest national parties after the 2019 election, each with 29 MEPs, were the German Christian Democrats and the British Brexit Party, which left the EP when the UK withdrew from the EU.

In order to gain recognition and resources, nationally elected MEPs must join a multi-national Party Group. The requirement that a Party Group has MEPs from at least one-quarter of member states reinforces the multi-national outlook of MEPs. The three largest Party Groups each have MEPs from more than 20 countries, whereas the Eurosceptic Groups have members from half the member states or less.

The policies of Party Groups are created by aggregating the views of MEPs who have won office by campaigning on a variety of national programmes. Within a Party Group, the national manifestos of their MEPs tend to show agreement in support of or opposition to further European integration, but limited agreement on other major dimensions of policy. This is particularly the case for the EP’s largest group, the People’s Party.

When there is a conflict between national public opinion and the policy of their Party Group, the MEPs in a Group tend to vote as a multi-national bloc rather than divide along national lines. Even though MEPs campaign by speaking to their voters in their national language, the great majority represent their constituents by working in a foreign language in a foreign country and voting for policies reflecting the consensus of a Party Group dominated by foreign parties.

Adding weight to democracy lite? Evaluating the democratic character of the EU by the standards of national governments justifies its description as light on democracy. This judgment encourages proposals to introduce reforms to give citizens more opportunities to make inputs to the EU’s policy process.

Since 1973 the European Commission has sponsored the Eurobarometer, a multi-national survey of public opinion about EU institutions and activities ( Its purpose is practical market research to identify means for promoting commitment to the European Union and its policies. Surveys thus ask questions about popular demand for EU services and how popular the EU is. Answers tend to show a plurality of positive responses; criticism is diluted because of the large number of respondents with no opinion about the European Union.

In response to pressures to give European citizens a direct say on EU policies, in 2012 the EU introduced what it called the Citizens’ Initiative. Citizens can petition the European Commission to consider an issue as long as it is within its existing powers. If the request is endorsed by a million citizens in at least seven member states, the Commission is obliged to publish a reasoned reply to the petition. The subtitle of a major academic review of this EU initiative is The Myth of a Citizens’ Union (Blockmans and Rusack 2019).

Party Groups sought to increase engagement with the EP electorate in 2014 by each nominating a Spitzenkandidat (German for leading candidate) for the presidency of the European Commission. Given linguistic and geographical obstacles, there was no campaign comparable to that for electing a national president. In 2019 the EPP was reduced to less than one-quarter of MEPs, but it remained the largest party. Its Spitzenkandidat, Manfred Weber, a German, had no experience of the Council as he had never held a

Making the whole of the EU a single constituency so that European citizens could vote in common for a transnational list of several dozen candidates has been mooted for decades (Duff 2018). The list would be required to have candidates and seek votes from a variety of countries, unlike the nationally elected MEPs that currently constitute the whole of the EP’s members. When raised by an EP committee in 2018, the idea of MEPs elected by a pan-European rather than a national constituency was rejected by a majority.

Formally linking national parliaments with the European Parliament has been proposed as another way to reduce the EP’s democratic deficit. However, most national parliaments already have committees to scrutinise policies and regulations coming from Brussels, and MPs can apply pressure on their national government to object to a particular policy in the European Council. To give national parliaments power to vote on EP measures raises the same problem as allocating seats: How would the votes of national parliaments be weighted?

In response to the EU almost doubling in size by admitting countries from Eastern Europe, in 2000 Romano Prodi, then the president of the European Commission, proposed measures to remove the ‘paralysis’ that he said was preventing EU decision-making, especially the right of a single member state to veto a proposal. Since Prodi’s proposals would have greatly limited the influence of member states, both large and small, they were rejected.

Prescriptions to strengthen the role of citizens in the EU policy process assume that its political problems are due to a lack of policy inputs. However, if decisions about EU policy decisions are viewed as technically complex, then giving power to technocrats rather than voters is a logical alternative. This has been done in the European Central Bank. However, making the EU more technocratic risks trouble if the promised benefits of expert policy-making are not delivered.

Proposals to add weight to the democracy-lite institutions of the EU face both structural and political obstacles. Gross disparities in population and economy between member states make it impossible to give European citizens the equality of representation that they enjoy in their national parliament. Population inequalities also prevent granting each member state a single vote in the European Council. Political opposition to reform comes from existing holders of power.


This paper is based on a chapter by Richard Rose in