A selection of experts answer a new question from Judy Dempsey on the foreign and security policy challenges shaping Europe’s role in the world.


Cornelius AdebahrNonresident associate at Carnegie Europe

The short answer is: nearly always. Unless constitutionally prescribed under specific circumstances, referenda are tools in the hands of elected politicians—or unelected ones, if they successfully press the government to put an issue to a vote. More often than not, those calling for a referendum end up like the narrator in Goethe’s The Sorcerer’s Apprentice, unable to control the “spirits that I’ve cited.” This is true from France’s 2005 referendum on the proposed European Constitution to the UK’s vote on EU membership and Colombia’s plebiscite on the peace agreement with the FARC guerrillas in 2016.

Tempting as it may sound to consult the people on a particular issue, the response is likely to be distorted by other factors: dislike of those in power, disinterest in the question on the ballot paper, dominance of a particular discourse. A referendum is less about solving a complex question (which it cannot) than about winning a domestic argument. As a telling sign, in the wake of the migration crisis, right-wing parties across Europe have come to favor elements of direct democracy that used to be the hobbyhorse of the Left.

The beauty of the system of representative democracy is that people can vote for a government to legislate based on clear procedures and don’t have to decide on individual policies. If there is a widespread lack of trust, as in Western democracies at the moment, asking citizens to take up the job of co-legislators is not the answer. Involving them more in existing processes is.


Rosa BalfourSenior fellow in the Europe Program at the German Marshall Fund of the United States

Little wisdom is to be found in the contemporary use of referenda. They are sold as tools of direct democracy to promote political participation among citizens who have turned away from politics. But in practice, they prize majoritarian conceptions of democracy at the expense of pluralist debate.

Ten years ago, political leaders would fudge unfavorable referendum results, such as the Irish people’s rejection of the Lisbon Treaty, making a mockery of this instrument as a measure of citizens’ preferences. Today, referenda have become political tools for other ends. The June 23 UK referendum on EU membership was about a major divide in the ruling Conservative Party. The upcoming constitutional referendum in Italy was construed as a plebiscite on the government but is becoming a means to oust it. The supposedly consultative referendum on April 6 in the Netherlands that rejected the EU-Ukraine Association Agreement is blocking government policy, even if only a minority of eligible voters expressed their opinions. After the October 2 referendum in Hungary on migrant quotas, which did not reach the required turnout threshold, Prime Minister Viktor Orbán is reiterating his illiberal policies.

The paradox is that the more governments lose referenda, the more they seem to call them. This leads to the suspicion that politicians use referenda merely to divest themselves of taking responsibility for political decisions. Rather than direct democracy, this use of referenda suggests a future of dictatorships of leaderless majorities.


Benedetta BarbisanAssociate professor of comparative constitutional law at the University of Macerata and fellow at the Max Planck Institute for Comparative Public Law and International Law

Referenda in Europe seem dangerous, especially for those announcing them. In the cases both of former British prime minister David Cameron with the June 23 Brexit vote and of Hungarian Prime Minister Viktor Orbán with the October 2 EU migrant quota referendum, it was as if the leaders unleashed a beast with the utter self-confidence of being able to tame it, only to be proved dramatically wrong.

Observers cannot know whether the constitutional referendum to be held in December in Italy will turn similarly fatal for the country’s prime minister, Matteo Renzi. His self-confidence was at such heights only a few weeks ago that he emphatically pledged to resign if defeated. Unfortunately, backpedaling when polls started showing a serious risk of debacle may have missed the mark.

At a time when anti-establishment sentiments haunt the entire continent (and no less the other side of the Atlantic), voters may take advantage of referenda to get rid of the leading class running the country and to express their yearning for the long-gone good old days. It is only in the best cases that people at the ballot box are moved by a critical analysis of the actual content of the question. In an era of market research and statistics, are voters fully aware of the difference between a referendum and a poll?


Federiga BindiSenior fellow at the School of Advanced International Studies at Johns Hopkins University, director of the Foreign Policy Initiative at the Institute for Women’s Policy Research, and D. German distinguished visiting chair at Appalachian State University

No, referenda are not dangerous; in fact, they are a precious instrument of direct democracy.

The problem is the inadequate or even inappropriate use of referenda. Former British prime minister David Cameron used the promise of a referendum on the UK’s EU membership to win the 2015 general election. Hungarian Prime Minister Viktor Orbán is using the October 2 referendum, which rejected the EU’s migrant resettlement scheme but failed to reach the turnout threshold, to continue his anti-EU rhetoric. Italian Prime Minister Matteo Renzi’s personalization of the forthcoming referendum on constitutional reform may prove fatal to him. Former Norwegian prime minister Gro Harlem Brundtland’s mistake in 1994, when she led the campaign to take Norway into the EU, was to put an existential and divisive issue to a vote too abruptly. The same thing happened in Colombia’s October 2 referendum on the peace agreement with the FARC guerrillas.

Many democratic leaders have fallen to the lures of referenda, often overestimating themselves—as in the cases of former French presidents from Charles de Gaulle, who held a vote to reform the senate in 1969, to François Mitterrand, who asked voters in 1992 to support the EU’s Maastricht Treaty, to Jacques Chirac, who called a referendum in 2005 on the proposed European Constitution. Few leaders have shown the strategic wisdom of former Spanish prime minister Felipe González, who slowly prepared Spain for a vote on NATO membership before actually calling a referendum in 1986.


Krzysztof Blusz and Paweł ZerkaVice president of WiseEuropa and head of the Foreign Policy Program at WiseEuropa

After recent frustrating results close to home (in the Netherlands on the EU-Ukraine Association Agreement, in Britain on EU membership, and in Hungary on the EU’s migrant resettlement scheme) and in distant places (in Colombia on a peace agreement with the FARC guerrillas), it may be tempting to dismiss the entire idea of referenda. In today’s context of heightened popular agitation, they seem to operate more as a risk than as a cure for democracy and its deficits.

There is a lot of demagoguery, populism, and political conformism behind the recent fashion for referenda. They have become low-hanging fruit for the burgeoning variety of political leaders who relish in positioning themselves as exponents of the common will while hurrying to put even the most strategic or complex issues to a popular plebiscite. All too often, referenda have pitched capricious public sentiment against well-informed decisions. They underwrite short-termism in public debate. The value of referenda in the domains of foreign policy, security, and defense is particularly doubtful.

The lesson to be learned is that any referendum should at the least be designed to avoid locking a country into an irreversible scenario. Allowing for the overdose and veneration of plebiscites does not guarantee democratic inclusion, education, or engagement in public debate. If governments cannot avoid referenda, they should use them as a democratic shortcut and a supplement to representative democracy—but never as its substitute.


Fraser CameronDirector the EU-Russia Center

Referenda are not dangerous as such, but much depends on the context in which they are held.

As inherently populist mechanisms, referenda have been used by demagogues throughout history to justify and extend their rule. Think of Hitler and most Communist regimes.

Referenda can be useful and legitimate to endorse a new or changed constitution. But again, the rules are important. A new constitution proposed by the Thai military government was endorsed in August 2016 despite severe restrictions on the opposition campaign.

In-or-out referenda on the EU are legitimate, but arguably there should be strict rules about the campaign to prevent the sorts of lies told in the run-up to the June 23 Brexit vote.

Another important issue is the demos for the referendum. Take the Dutch referendum in April 2016 on the EU Association Agreement with Ukraine. The demos consisted of the 500 million citizens of the EU, not just the 17 million citizens of the Netherlands. So there should have been an EU-wide referendum or none at all. An alternative could be that EU-wide agreements enter into force only with the consent of member states representing 80 percent of the total EU population.

The real danger of referenda is that they are often used by politicians seeking short-term political gains. In Europe, there is a tradition of parliamentary democracy. Let’s stick to it.


Thomas de WaalSenior associate at Carnegie Europe

Democracy is a process, not a binary yes-no decision. Referenda should be limited to very specific topics—the building of an airport, some social issues—on which the question is clearly defined.

The June 23 Brexit vote, the April 6 Dutch referendum on the EU Association Agreement with Ukraine, the October 2 Colombian vote on the peace agreement with the FARC guerrillas: all of these invited voters to give a simple verdict on arrangements of enormous complexity that had taken years to negotiate.

In some cases, the outcome may harm those same voters who voted for it. Colombia risks new violence, and the regions of the UK that voted Leave are those that will see the greatest economic damage from exiting the EU. That is why most democratic countries have representative democracy: a vote for those deemed professionally qualified to make decisions on citizens’ behalf. The alternative ends up becoming government by opinion poll, with the public giving snap judgments on political issues in the same way that they vote in Britain’s Got Talent.


Catherine FieschiDirector of Counterpoint

Referenda embody the worst kind of democracy: one in which majorities lord it over minorities in ways that are at best unfair and at worst oppressive. In diverse societies, this is a problem. But it may actually be a lesser evil.

Bigger problems arise when referenda are used in particularly contentious cases that divide the electorate into large, relatively balanced blocks: then the losing side is a big minority that cannot be easily discounted. Political instruments that ride roughshod over a large proportion of the population are neither democratic nor efficient. The June 23 UK referendum on EU membership is a case in point—leaving a trail of chaos, anger, and contempt in its wake.

This is by far the most dangerous consequence of referenda: by absolving leaders of decisionmaking, these votes also cancel out politicians’ function as political and social lightning conductors, leaving citizens to face each other without much mediation from institutions. In representative democracies, politicians and policymakers govern on citizens’ behalf—but, crucially, they also absorb the force of people’s discontent to avoid its dispersion in civil and political society. Referenda shut the door on that possibility: politics becomes a showdown.


István HegedűsChairman of the Hungarian Europe Society

Yes, referenda can be dangerous, even for populists. Hungarian Prime Minister Viktor Orbán’s government could not convince enough citizens to vote in its October 2 anti-refugees, anti-Brussels referendum to make the ballot valid. Populist politicians always attack mainstream democrats for having lost contact with social realities and ordinary people. This time, ironically, a pointless populist question about the “obligatory resettlement” of non-Hungarians, allegedly forced on Hungary by the EU, failed to be attractive to a majority of the electorate. The nationwide inhuman communication campaign did not help either, even though the government spent more on billboards and ads, which mixed the refugee crisis with terrorism, than the Leave and Remain camps combined before the June 23 UK referendum on EU membership.

Still, for nationalist and illiberal political forces in opposition, this tool of direct democracy is an attractive weapon against the status quo and the liberal institutional setup. In other cases, referenda can counterbalance and correct shortsighted decisions by parliamentary majorities as well as local authorities.

At the European level, however, national referenda raise a mathematical problem. In the Hungarian case, the senseless referendum had no legal consequences for the EU. But what would happen if all EU member states asked their national citizens to approve or refuse decisions made by the European institutions? Given the current mood of Europeans, deadlock would be inevitable.


Hanspeter KriesiStein Rokkan chair of comparative politics in the Department of Political and Social Sciences at the European University Institute

Some are, some are not. Whether they are depends above all on how they are organized.

Plebiscitary, top-down referenda that are called by political elites to serve their own purposes—such as those called by former French president Charles de Gaulle on constitutional reforms, former British prime minister David Cameron on the UK’s EU membership, or Hungarian Prime Minister Viktor Orbán on the EU’s migrant quotas—are dangerous. They are easily manipulated by the wording of the question, the organization and financing of the campaign, and the interpretation of the outcome. Such votes introduce great uncertainty into the political process.

By contrast, constitutionally prescribed referenda or those initiated by bottom-up mobilization as a routine part of the political process, as in Switzerland or the United States, are not dangerous. These votes tend to fulfill an integrative, legitimating function by improving political decisionmaking, refining the outcomes of social and economic policies, and contributing to citizens’ general satisfaction—as studies of referenda in Switzerland have amply shown.


Enrico LettaPresident of the Jacques Delors Institute, dean of the Paris School of International Affairs at Sciences Po, and former prime minister of Italy

In the search for a policy based on rapidity, simplification, and spectacle, referenda have become very fashionable tools. Do they also bring efficiency and balance to the decisionmaking process? Do they help democracies make choices that the people will not regret? I don’t think so.

Referenda, inserted into institutional systems of representative democracy, are often a false lead, a shortcut that gives a superficial impression of helping democracy, which is in fact weakened. They are too strongly linked to current emotions and too strongly influenced by media events occurring at the time of the vote.

But what normally makes a referendum a double-edged sword is the decisive importance of how the question is asked. The framing is always in the hands of whoever proposes the referendum; and even if the proponents are beaten, as in recent cases, the fact that the question is posed by a political actor such as a government or political movement makes the exercise unbalanced from the start. And the final result always depends on this initial imbalance.


Denis MacShaneFormer UK minister for Europe

Referenda were described by Margaret Thatcher as a “device of dictators and demagogues.” She was quoting Clement Attlee, and it is hard to disagree with Britain’s two greatest postwar prime ministers. The last major European nation to quit an international treaty organization (which is what the EU is) after a referendum was Germany in November 1933, when Germans voted to leave the League of Nations. Although Hitler was in power, academics agree the referendum was fairly run and reflected the rise of nationalist-populist feeling against membership of a supranational rules-based organization.

Sensibly, Germany banned referenda after 1945, and the United States does not hold national plebiscites. They work in fully federal Switzerland as a means to hold together conflicting cantons, religions, economic interests, and language groups. But they are a failure of politics based on rational and representative democracy. They open the door to rich men who can pay for the preparatory work to secure the holding of a referendum. The British plebiscite on EU membership on June 23 was marked by an unprecedented campaign of hate against foreigners and—it is now widely accepted—was run on lies that were not challenged by the BBC or other media.

If Republican candidate Donald Trump becomes U.S. president in November, will he try to introduce referenda into national political decisionmaking in the United States?


Marc PieriniVisiting scholar at Carnegie Europe

A referendum is part of the democratic toolbox, but in the current wave of populism, it can easily be hijacked to serve specific political aims beyond democracy itself. In Europe, referenda have increasingly been used to obtain emotional answers to emotional questions as part of a political strategy.

The April 6 referendum in the Netherlands on an EU-Ukraine agreement—a rather obscure topic for a direct ballot—was used to illustrate the existence of a large anti-EU feeling among Dutch citizens. Similarly, the campaign leading up to the June 23 Brexit vote was based largely on false assumptions on the part of the Leave camp, and the result left the UK wondering whether it had made a sensible choice. The October 2 Hungarian referendum on refugee policy, although it failed to reach its own turnout threshold, was and remains an instrument of Prime Minister Viktor Orbán’s populist anti-EU stance.

In France, the National Front has proposed an in-or-out referendum on EU membership similar to Britain’s, while another presidential hopeful has suggested holding referenda on topics related to nationality and suspected Islamists. The goal is obvious: flatter the anti-EU and xenophobic sentiments of part of the public, rather than explain and convince.

Asking a question on a narrow, technical, and emotional subject can lead to distorted answers. Western democracies should beware of putting too often simplistic questions to a direct vote. They have a lot to lose in the medium and long term.


Gwendolyn SasseNonresident associate at Carnegie Europe and director of the Center for East European and International Studies in Berlin

Classifying referenda as generally dangerous is an approach contemporary democracies can ill afford. At a time when even long-existing democracies are confronted with a weakening of mainstream political parties and low electoral turnout, especially at the local level, referenda offer the potential to reconnect the electorate with policy outcomes.

However, referenda bear political risks, not least because they can develop their own dynamics. Two aspects are particularly significant: the rules governing a referendum (the type of issue that can be put to a vote, the level of government at which the plebiscite is held, and the mechanics of the ballot) and the wider political context in which a referendum takes place.

Referenda are not a panacea to fix the problem of democratic legitimacy. The specification of referendum rules, especially the turnout threshold, requires careful consideration. Ultimately, the wider political context is critical. As the June 23 Brexit vote has demonstrated, against a backdrop characterized by a lack of public discussion of the issue at stake, a referendum campaign is unlikely to fill the information vacuum and offers ample scope for political manipulation. This experience, however, should not invalidate the potential of referenda as one part of the democratic toolbox.


Joseph C. SternbergColumnist and editorial-page editor of the Wall Street Journal’s European edition

Of course referenda are dangerous, because they’re so ineffective at settling vexing debates. A plebiscite tells you how some voters feel about a particular question at a particular time. But it’s silent on how voters prioritize the question on the ballot relative to other concerns. And it’s silent on what voters think about the wide range of smaller policy choices required to implement most referendum results.

Take Brexit. A relatively narrow majority of British voters on June 23 wanted to leave the European Union. But for decades, including in 2015, voters returned members of parliament who felt otherwise, suggesting that relative to other matters, Brits weren’t actually too bothered about the EU. And the referendum offered no judgment on specific problems related to immigration, trading relations, and the like that now roil politics, while the 48 percent who voted Remain seethe.

Hence the danger. Referenda short-circuit the parliamentary method for setting priorities and limit adaptability by binding voters to one decision made on one day. And they curtail a polity’s ability to account for minority concerns via political trade-offs to preserve a sense of fair play rather than let the winner take all. At a time of rising political disorder, representative democracy needs to work better, not be circumvented entirely.


Stephen SzaboExecutive director of the Transatlantic Academy

Referenda on foreign policy questions are dangerous because of the low levels of public knowledge about issues that are far removed from the day-to-day experiences of the vast majority of the population. These votes tend to have low turnouts and bring out the most committed parts of the public on the issue involved. As is often the case, the public follows elite cues on foreign policy, and questions in this area are low on the electoral agenda in almost all elections.

Irish statesman Edmund Burke wrote eloquently on the responsibility of elected representatives to weigh up the interests of the electorate on issues on which competing priorities and values need to be balanced. If voters are unhappy with the results, they should replace their representatives rather than be asked to weigh in on issues beyond their own levels of competence.