They are the national politician’s weapon of last resort against more European integration. When the Greek government feels cornered by the demands of its creditors, when British Prime Minister David Cameron faces mounting Euroskepticism in his party, when the French government needs to respond to popular fears concerning further EU enlargement, national referenda are scheduled—or at least threatened.

As trust in Europe’s mainstream parties has declined and populist movements have grown in many EU countries, demands for referenda to block or reverse integration steps have proliferated. So far, the political elites have resisted many of these demands, but the pressure for more direct democracy is clearly growing.

Fear of failed referenda has stifled EU leaders’ readiness to reform the union’s founding treaties and adjust the EU to new challenges. As the ultimate democratic expression of the people’s will, these votes are important features of democratic life, but they seem increasingly difficult to reconcile with progress on European integration. If current trends continue, the EU’s capacity to reform will diminish even further.

There is no easy solution to this problem. However, making the EU’s normal decisionmaking process more transparent and participatory could help restore confidence in the system. If citizens felt more involved in and consulted on EU policymaking, the pressure for referenda might ease up.

The People’s Will vs the Sausage Machine

In purely aesthetic terms, referenda are high points of democracy. Like in ancient Athens or today’s Swiss towns, the real sovereign—the people—decides on a matter directly without mediation by institutions or political parties. No secret deals in smoke-filled rooms, no lobbyism, no horse-trading; every citizen of voting age participates on an equal basis. At the end, there is a clear result without ifs or buts. No other type of democratic decisionmaking has similar dignity and legitimacy.

The EU’s normal method of reaching decisions couldn’t be more different. Like in a gigantic sausage-making machine, different viewpoints are mixed together and kneaded through, their edges removed, their contradictions fudged. All this is then tweaked and packaged until in the end it has become sufficiently obscure who has won and lost, and every minister can declare victory in a national press conference. As an instrument for accommodating the divergent interests of a large number of very different member states, this system has worked well for many years—but at a considerable price in terms of popular participation and transparency.

Since the French people were first called on to vote in 1972 on enlarging the European Community (the EU’s predecessor), these two types of European decisionmaking have coexisted. So far, 54 referenda on issues of European integration have been held. Apart from the votes organized by EU member states or acceding countries, this figure includes eleven referenda in Switzerland and Liechtenstein on their relationships with the EU as well as two Norwegian votes against EU membership.

Nineteen referenda dealt with the question of accession. While the initial community was founded in the 1950s without any consultation of the people, it later became customary that at the end of long accession negotiations, the population of the acceding state would have the final say on membership. In view of the importance of this decision, these referenda are generally perceived as good democratic practice. In the vast majority of cases, the outcome has been in favor of accession.

Referenda on leaving are much rarer. The population of Greenland, which entered the community as a dependency of Denmark, voted in 1982 to leave. The UK joined without a referendum in 1973 but two years later held a referendum on staying in. Four decades on, the question of the UK’s place in Europe is still not resolved. The current British government has committed itself to holding a further in-or-out referendum, at the latest in 2017, but more likely in 2016.

Other popular votes have related to specific policy issues, such as participation in the eurozone (Denmark and Sweden) or support for a bailout program (Greece). The next referendum of this type will be Denmark’s decision on December 3, 2015, about whether to take part in more aspects of the EU’s judicial cooperation, from which it is currently exempt. Following a public initiative, the Netherlands will organize a nonbinding referendum in 2016 on the ratification of the EU’s Association Agreement with Ukraine. This is the first time that an important foreign policy issue (apart from enlargement) will be submitted to a referendum.

The Minefield of Treaty Change

Most of the other EU referenda have concerned the approval of treaty reforms that aim to develop the structure and the policies of the union. And it is these referenda in particular that worry the proponents of further European integration.

Fear of referenda has prompted governments to avoid treaty change by resorting to secondary legislation—laws other than treaties—or creative reinterpretation of the treaties. Thus, paradoxically, the increasing demand for direct democracy has resulted in more technocratic fixes and even less proper democratic scrutiny, aggravating citizens’ already-widespread sense of alienation.

Reforming the EU’s founding treaties requires a so-called double lock: the agreement of all EU governments on a new treaty text, followed by ratification by all member states through either a parliamentary decision or a referendum. A single no from one out of 28 member states can kill an important project. Thus the EU’s constitutional treaty was sunk by negative votes in France and the Netherlands in 2005. Negative referendum results in Denmark and Ireland delayed the entry into force of the Treaties of Maastricht, Nice, and Lisbon.

Referenda on treaty reform highlight the tension between democracy on the national and on the European levels. Within a national context, it can appear a democratic necessity that the people should decide on developments of EU structures and policies that will have important implications for the country. However, from the point of view of the union as a whole with its more than 500 million people, it is questionable whether the failure of a treaty resulting from the negative vote of a small majority in Ireland, for instance, with its population of 4.6 million, can be considered a triumph of democracy.

In the past when the people said no to a new EU treaty, this was not the end of the story. Soon after the outcome of a referendum was known, the EU’s other type of decisionmaking process, the compromise machine, was switched on again and set to work on the consequences. And that often led to results that called the theoretical finality of referendum decisions into question. Denmark and Ireland found out that they would have to vote again if they didn’t want to forgo important benefits of integration. Opt-outs and other mainly symbolic concessions made it easier to swallow the bitter pill of having to go back to the polls, and in fact, second votes on treaty change have so far always been positive. Even the sinking of the constitutional treaty by the French and Dutch people was not final. Two years later, some 90 percent of the document’s provisions resurfaced in the form of the Lisbon Treaty, which entered into force through parliamentary procedures in all member states except Ireland, which held two referenda on this issue.

The real paralyzing effect of referenda therefore lies—at least so far—not so much in their actual record in bringing down reform projects as in their psychological impact on EU governments. Nothing dampens the willingness of European governments for reform more than the fear of failed referenda.

Between 2008 and 2012, EU leaders enacted far-reaching measures to shore up the bloc’s monetary union. New rules for fiscal discipline, bailout funds, and a eurozone banking union involved considerably deeper structural change than several earlier treaty revisions. However, EU leaders chose to stretch and tweak the existing treaty framework rather than amend the treaties themselves. Just two measures—the fiscal compact aimed at fostering budgetary discipline and the European Stability Mechanism designed to provide financial assistance—were adopted in the form of intergovernmental agreements outside the EU treaty framework.

Of course, there are limits to the extent to which the existing treaty provisions can be stretched. It is broadly accepted that the further steps needed to make monetary union sustainable or to equip the EU to deal effectively with the implications of the current migration crisis cannot be accommodated within the present treaty framework. But the discussion of a more comprehensive reform project keeps being postponed because the risks of a failed ratification process are deemed too high.

An Ever More Popular Instrument

Why, then, are referenda on EU issues increasingly in vogue these days? To some extent, this tendency reflects a widespread disillusionment with representative democracy. Due to globalization, governments are less and less capable of shaping economic and political developments in their countries. Hence, they frequently fail to deliver on their promises and consequently lose the trust of citizens. In this situation, support for direct democracy has grown in many advanced democracies. Out of fear of being held accountable for unpopular decisions, governments often quite willingly give in to such demands.

The rise of Euroskepticism too is behind many of the current calls for referenda on issues relating to European integration. For people for whom integration has developed much too fast, a referendum is the obvious emergency brake. The fact that the UK, with very little past experience of direct democracy, has adopted the most extensive referendum lock on EU issues is a case in point. In the face of growing Euroskepticism in the Conservative Party, the Cameron government in 2011 passed legislation making any further transfer of powers to the EU subject to a referendum.

The traditional pro-European elite remains skeptical of the use of referenda to make EU decisions. Members of the elite highlight the great complexity of many EU decisions, which, they claim, makes it tough for nonexpert citizens to form an opinion. They point out that well-financed single-issue interest groups can hijack referendum campaigns in ways that are hardly possible in the case of normal elections, in which a broad variety of issues is usually at stake. Also, the process leading up to a referendum is often influenced by factors that have nothing to do with the matter at hand. External or internal political turbulence, the state of the economy, or the popularity of the current government can impact the result. Referenda always produce a response, but not always to the question that was actually asked.

While not without merits, such arguments are beside the point. Once the referendum genie is out of the bottle, it is very hard to put it back in. It is difficult for politicians to argue that the people are not wise enough to make a decision. Notwithstanding the great variation in constitutional and political systems among member states, there is a bandwagon effect. If a referendum is granted in one country on one issue, it becomes harder to deny it in another country or on another issue.

No Way Around Referenda?

If the current proliferation of referenda continues, is it plausible that the EU will be able to deal with complications ensuing from negative results in the future? The task has become much more difficult of late. The EU is still grappling with the consequences of the 2007–2008 global financial crisis, which has reduced citizens’ confidence in governments generally and in the EU particularly. The refugee crisis has further emotionalized the European debate and reinforced centrifugal tendencies. And all this comes at a time when traditionally pro-EU mainstream parties are losing support while populist and often Euroskeptic movements are gaining ground.

The risk of negative referendum results has increased, and the task of reversing such outcomes has become more complicated. Well aware of this situation, EU governments are likely to continue to shy away from reforms that involve treaty change.

The double lock on treaty change—unanimity among governments on the substance of the reform plus ratification by each member state—certainly constitutes a problematic constraint on the EU’s capacity to adjust to new challenges. The increased number of member states as well as the tendency to choose a referendum as the means of national ratification threatens to turn this constraint into a full-fledged blockage.

Supporters of dynamic EU development therefore advocate more flexible ways of deciding on constitutional reform. Most of these proposals would eliminate the requirement of ratification by every single member state. Some reformers also suggest a pan-EU referendum on treaty change. Treaty amendments would pass only if both a majority of member states and a majority of the total voting population voted in favor of the changes.

The trouble with such new methods of constitutional reform is that legally, they would first have to be adopted in the traditional way of unanimity among governments and ratification by all. As these reforms would entail a massive curtailment of national sovereignty, it is hard to see this change coming about in the foreseeable future.

Aware of this obstacle, some advocates of deeper integration call for a rupture with the existing legal order. They suggest a constitutional assembly to come up with a new EU constitution, which would then be put to the vote in a pan-European referendum. Only states in which the constitution is supported by a majority of the population would participate in the new, refounded union.

This scheme assumes that the leaders of a large number of EU countries would be ready for a historic gamble with unforeseeable consequences. This is highly implausible. Fighting the spread of Euroskepticism with revolutionary federalization may have romantic appeal, but it seems quite far-fetched in terms of practical politics.

There is thus no constitutional silver bullet to overcome the referendum lock on EU reform. Any other grand scheme for a solution involving treaty change would probably also fall victim to failed referenda.

What is more, the EU’s traditional method of turning political disagreements into technical issues that can be resolved through expert negotiations behind closed doors has become unsustainable. As a result of the euro crisis as well as through its work on migration and home affairs, the EU has become involved in areas of great salience for the public. The current controversy about TTIP, the proposed Transatlantic Trade and Investment Partnership, also shows that issues that earlier were left to European Commission technocrats and remained largely outside the public eye have now become highly politicized.

The more or less successful coexistence between the established, elite-driven, technocratic process of EU decisionmaking and the sporadic, direct expression of the people’s will through referenda is therefore coming to an end. In the more contested and politicized EU of today, the tensions between the traditional method and the demand for direct democracy will grow. If EU leaders cling to their usual way of doing business, they will increasingly face the blockage of referenda. Further technocratic fixes and legal ingenuity may win some time, but eventually the day of reckoning will come.

Toward More Participatory Decisionmaking

The widespread demands for more direct democracy in the EU show that many citizens strongly desire to be more systematically involved in EU decisionmaking. If EU leaders wish to avoid a proliferation of referenda, which could easily lead to paralysis, then they will have to address this popular wish in another way. Making the EU’s normal decisionmaking process more transparent, participative, and democratic offers the most plausible alternative. This can be achieved through an incremental process involving an array of measures on different levels and in different areas. Considerable progress can be reached without treaty change.

Cyberspace has unused potential for better connecting the EU to the citizens. At present, there is plenty of information about EU matters around, but it is often hard to access and understand for the nonexpert. EU institutions can use the Internet to broaden access to information and make it more user-friendly. Every citizen should have the possibility to follow the decisionmaking process on issues of interest.

Modern means of communication should also be used to broaden the space for interaction. Existing mechanisms such as the European Citizens’ Initiative, introduced by the Lisbon Treaty to allow 1 million EU citizens to call on the European Commission to launch a legislative proposal, or the right of petition are either too cumbersome or too unknown to give a sense of real involvement. EU institutions should open up existing consultation processes to more participation from civil society. Innovative deliberative mechanisms should allow and encourage citizens’ direct participation. Demands for referenda will have less support if important issues are thoroughly debated through other political processes.

As the divisions between European and national politics are breaking down, there is a need to better integrate these two spheres. EU-level politicians are at one remove from voters, so national leaders have a key role in reconnecting the EU with the public. National parliaments should therefore be more systematically involved in EU decisionmaking. More member states should follow the examples of Denmark, Finland, and Germany, whose parliamentary EU scrutiny committees already play a powerful role. Members of the European Commission and of the European Parliament should participate in the debates of national parliaments on EU issues. EU institutions should also systematically reach out to regional and municipal authorities, which on the whole enjoy greater trust from the citizens.

These are just some of the measures the EU could use to turn itself into a space for genuine transnational public debate. If, as a result, citizens felt more involved and consulted, their confidence in the system would rise again and the pressure to use referenda as defensive weapons of last resort would diminish.