What If Europe Held an Election and No One Cared? by David Schleicher
[David Schleicher, an Associate Professor of Law at George Mason University School of Law, describes his recently published article What If Europe Held an Election and No One Cared?]
So, to start, I would like to thank my editors at the Harvard International Law Journal and the good people here at Opinio Juris for providing this forum. And I would particularly like to thank Erin Delaney and Samuel Issacharoff for writing a response to my paper. I’m looking forward to hearing your responses as well.
My paper, What If Europe Held an Election and No One Cared?, examines the difficulties the European Union (EU) has had in introducing direct democratic representation into its law-making process. In so doing, it provides an explanation for why multi-level political systems frequently feature a common type of democratic failure. Elections at non-national levels of government (ranging from the pan-European elections I discuss in the paper to state and city legislative elections in the U.S.) regularly fail to provide much in the way of democratic accountability because voters treat them as a referendum on national political figures while paying little attention to the effect they will have on local, or supra-national, public policy. In the paper, I argue this is a result of “mismatch,” or an election law system that causes (or doesn’t discourage) a lack of fit between the governmental level at which political party systems are organized and where elections are held. Looking at the EU provides a window into this endemic problem for multi-level democracies.
So, some background: In 1979, the EU introduced direct elections for European Parliament (EP), and successive treaties since then have turned the EP into a central part of the EU’s law-making process. After the passage of the Lisbon Treaty in 2009, the EP became a near co-equal to the other two branches of the EU’s law-making apparatus: the Commission, the bureaucratic heart of the organization which serves as both executive and the entity with the power to propose all legislation, and the Council, where Member States vote directly.
The decision (or rather decisions) to make the EP more powerful has been in response to a popular perception that the EU suffers from a “democratic deficit,” that its power has grown at the expense of domestic governments but the people of Europe have little control over its decisions. In the institutional logic of the EU’s separation of powers, the EP thus has a key role to play. It is supposed to be a democratic and European check and balance on the bureaucratic and national forces represented in the Commission and Council. The EP is the institution through which European voters can comment directly on European policy.
However, there is a problem – voters simply do not use EP elections to say anything about the EU. The clearest finding in contemporary European political science has been that the results of EP elections do not turn on anything that happens in Strasbourg or Brussels. EP elections are “second order,” contested by domestic parties and with results that turn almost exclusively on how popular domestic Prime Ministers are (the only exception is that small parties and anti-European parties do slightly better in EP elections than they do in domestic ones.) This can be put starkly: nothing a Member of European Parliament (MEP) has ever said, and nothing one has ever done, has ever effected an EP election. Turnout has been falling consistently since the first direct election in 1979, and the campaigns are more farcical than substantive. EP elections end up being like midterm opinion polls on domestic governments, except with real consequences, as the EP now is a major policy-maker on the European stage.
So what’s going on? The problem of EP elections is similar to a problem we see in American local and state elections, a problem I call in the paper “mismatch.” Where an election law system either encourages or does not discourage party systems developed at the national level to appear on ballots at another level of government, under-informed voters will use party preferences developed at the national level when voting at the local or state level as long as there is any correlation between their policy preferences at the two levels. As voters have little specific knowledge about individual politicians or what is going in most legislatures, this is perfectly rational behavior – they vote based on what they know, and they know national parties. As I have shown previously, unless parties develop locally differentiated sub-brands (more on this in a second), voters’ impressions of the President and Congress determine local election results even though there is substantial evidence that knowing a politician’s national party membership tells us very little about her preferences about local policy. As a result, local elections do very little to translate voter preferences about local issues into governmental policy and local officials are not particularly accountable for poor performance.
The EP works the same way. Voters see national parties on their ballots – Labor and Conservative in Britain, the UMP and the PS in France, etc. Because voter affiliations with national parties are strong and national parties appear on the ballot, voters rely on their feelings about national parties and politicians when voting for EP. The result is that EP elections are second-order. But why don’t national parties – particularly ones that are unpopular in a given election – try to develop platforms on EU issues that appeal to voters beyond their usual supporters, the same way that candidates for Governor are sometimes capable of using state-specific issues to disassociate themselves from unpopular national parties. In American local and state elections, election laws, like laws making party switching between elections difficult and requiring the use of popular primaries to selection candidates, make the development of sub-national party brands difficult, dooming all but candidates for the most prominent offices to election returns that rise and fall with their national party. In Europe, it is the structure of elections that ensures that results will be second-order. As no one national party actually governs the EP (no one country has more than 13% of the seats), voters cannot develop retrospective evaluations of how the parties actually perform on EU issues. This makes party differentiation on European issues difficult.
Political scientists have shown that under-informed voters have little ability to translate observations or preferences about politics into votes without party labels that provide a simple heuristic guide for voters to the ideological positions and past performance of their candidates. Without parties contesting EP elections along EU issues, EP elections end up not being about the EP or EU at all.
But what can be done? “Electoral engineers” have developed a number of tools for shaping partisan competition in transitional democracies to achieve certain constitutional results, and the EU should borrow from this tool kit.
Specifically, the EU should use “distribution requirements” to require the development of pan-national party brands. Currently, MEPs join together in like-minded coalitions – e.g. members elected from the Labor Party in Britain and the PS in France both join the Party of European Socialists, members from the UMP and Christian Democrats from Germany join the European People’s Party. These coalitions – “Euro Parties” – have been given legal identity in the EU, have big EU-funded budgets, and engage in highly effective whip operations inside the EP. If the EU adopted a distribution requirement of a type used in Nigeria and several other countries, these Euro Parties would appear on the ballot. The EU could require that no party that gets less than say, 3% of the vote in 7 European countries, gets any representation in EP elections at all. (Voting thresholds – rules that a party has to get above a certain percentage to win seats – are common in proportional representation elections and this would just modify the content of the threshold.)
This modest change would have a big effect. No major party competes in national elections or EP elections in more than one country. In order to win seats, national parties likely would run under the banner of the Euro Party with which they are affiliated. Voters would see Euro Parties on the ballot and would have the ability to develop an understanding of what these Euro Parties stand for and how they have performed. Further, having parties run under Euro Party banners would limit the ability of candidates to run beggar-thy-neighbor campaigns, as a Euro Party candidate from France who promised, say, to end an EU policy that was good for the British would end up dragging down his party’s ticket across the Channel.
One nice characteristic of this proposal is that it is tailored to the problem. It would have no effect on other parts of the EU system and would not undermine domestic parties in any direct way, even their control over candidate selection or campaigning during EP elections. It would simply inform voters in EP elections about what they were voting for. However, given what we know about how under-informed voters actually behave, this might be enough to produce real democratic gains. And if it worked, the EP would be able to its proper role in the EU’s separation of powers system.
More generally, thinking about what might make EP elections work should cast some light on how we might think about problems in elections in multi-level democratic systems generally. Mismatch problems are common in sub-national as well as supra-national elections. State legislative elections turn on Presidential popularity, county executives lose elections because Congress passed an unpopular bill, etc. This is a major problem for theories of federalism, and for constitutional allocations of power more generally. The specific type of solution I propose here may not be useful in solving these types of problems, but something of a similar flavor might. The problems of EP elections provide a dramatic instance of mismatch, and only by addressing it – either in the EU or in the U.S. – will our actual practice of democracy fit the theoretical models dreamed up by our constitutional framers and institutional designers.