You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.

The Case Against Referendums: From Greece to California, They Always End Up Undermining Democracy

In calling for a referendum on Greece’s bailout plan, Prime Minister George Papandreou has, it could be said, embraced one of his country’s oldest political traditions: direct democracy. The idea that the citizens of a state should all cast votes to decide matters of common interest was arguably born within an easy walk of his Athens office, some two and a half millennia ago.

Of course, referendums have remained a part of democratic politics into the modern era, with a formal place in the constitutions of many countries and regions, from France to Australia. In the United States, their use goes back to the town meetings of colonial New England, and they still have a crucial role in the politics of several states, notably California.

But the use of referendums as a political procedure has always been tense, even problematic, for democracies—and that is no less true of Greece’s planned vote in January. Papandreou says he is paying deference to the unmediated will of the Greek people, and so presents the vote as the very essence of democracy. Yet in fact, referendums have most often done more to weaken democratic institutions than to strengthen them, and this new example is no exception.


THE STORY BEGINS with a wholly obvious point: Modern states are far too large and complex for direct democracy. Since it would be hugely impractical for the people, as a whole, to decide on everything from the size of foreign aid budgets to new environmental regulations, they delegate the business of government to elected representatives.

We tend to forget, however, how very strange and unnatural this practice once seemed. Why should voters trust an ever-changing group of people, most of whom they themselves have not elected, most of whom they have never met, and with whom they may have nothing in common, to make decisions on matters that intimately affect their livelihood? Surely it makes much more sense (it was once thought) to entrust government to people you are bound to through ties of family, or class, or caste, or whom you believe God has chosen to rule.

As late as the eighteenth century, some prominent political philosophers rejected the very principle of representative government. Jean-Jacques Rousseau, for instance, mocked the English for believing that an elected parliament made them free. They were free only at the actual moment they voted, he insisted. As soon as they surrendered their political authority to their representatives, they lapsed into a state of slavery. Even in the midst of democratic revolutions, representative democracy remained an object of massive criticism and suspicion. In the 1790’s, France’s radical sans-culottes demanded that national representatives rigidly follow instructions provided by local electoral assemblies meeting in permanent session. When certain deputies to the National Convention aroused their ire, the sans-culottes stormed its meeting hall and successfully demanded their expulsion.

It does not help matters, of course, that in practice bodies of elected representatives so often seem to devolve into corrupt, complacent and long-lasting oligarchies. Anger at the shenanigans of the political class has helped keep the old suspicions alive right down to the present day, and has led, in democracies across the world, to countless institutional schemes designed to keep elected representatives in check: “imperative mandates” (detailed orders for how to vote in parliament, drawn up and approved by constituents); term limits; making the job part-time; judicial oversight; etc. The single most popular such scheme, however, has been the referendum. The logic, after all, seems unassailable. Who better to retain authority over the people’s representatives than the people themselves?

True, not all referendums fit this description. During the past two hundred years or so, two basic categories of referendum have developed. The first involves fundamental questions of sovereignty. In many countries, constitutional amendments require public approval by referendum. And following Woodrow Wilson’s formulation of the principle of national self-determination, the Versailles peace conference after World War I prescribed “plebiscites” for areas of contested nationality within Europe (the terms “plebiscite” and “referendum” overlap in practice). More recent examples include the unsuccessful referendum on independence in Quebec in 1995, and the successful one in East Timor four years later. And arguably, referendums of this sort are entirely appropriate. They represent instances when sovereign power, always ultimately held by the people, but mediated by constitutional structures, temporarily reverts to the people directly, so that they can modify or replace these structures.

But then there is the sort of referendum that addresses questions normally dealt with by elected legislators. Here, the example most familiar to Americans is the California ballot initiative, introduced exactly a hundred years ago by progressives seeking to wrest power away from corrupt political machines. Many of the most famous California initiatives have had the formal status of constitutional amendments, but often the matters at hand arguably belonged to the category of ordinary legislation: for instance 1978’s Proposition 13, which  placed severe limits on property tax rates, and 2008’s Proposition 8, which banned gay marriage. And even more often, Californian voters have had to vote on such mundane questions as whether to issue a new state bond, or exempt candy from sales tax, or impose an additional tax on cigarettes to fund cancer research.

It is certainly tempting to salute this second form of referendum as a means of checking the seamy practices that too often infect modern representative systems. But however much the designers of referendums claim to be acting in the name of democratic reform, their actions usually end up undercutting democratic institutions. This tendency isn’t merely incidental—it’s unavoidable given how referendums work. First, they take relatively technical issues away from legislators who have the time and expertise to deal with them, and give them to voters who do not.

Second, particularly when they come dressed up as constitutional measures, they tie the hands of legislators in potentially destructive ways. Take California, for instance. Without the ballot initiative procedure, voters incensed about high property tax rates in 1978 might have voted for legislators who promised tax relief. Instead, they approved the radical Proposition 13, which ever since has hamstrung attempts to generate sufficient state revenue, and contributed as much as any other factor to California’s current, destructive budget crisis (which, not so coincidentally, now poses something of a Greek-style threat to our own economy).

Third, and in some ways most importantly, referendums tarnish the legitimacy of legislators by subjecting their work to direct popular veto, and therefore casting it as a less genuine expression of popular sovereignty—despite the fact that the routine functioning of a democratic constitution is the most important expression of this sovereignty.

This last effect is even worse when legislators themselves propose referendums. In theory, again, the intentions are entirely honorable, and fit in both with ancient tradition and progressive policy. In practice, however, legislators in this position are often following another well-known parliamentary tradition, namely trying to squirm away from difficult decisions. But if a legislator asks for a popular vote to decide a measure, when a parliamentary vote would suffice, he or she is calling the parliament’s democratic authority into question.

Not surprisingly, referendums have often appealed to demagogic, authoritarian politicians. The man who probably did most to popularize their use in modern Europe was Napoleon Bonaparte, who used them to legitimize his seizure of power, and his transformation into an emperor. (He won some of them, but cheated anyway to increase his margin of victory.) And it is no coincidence that France today gives a more prominent constitutional role to referendums than nearly every other western democracy. Its present constitution was tailor-written for an imperious President who had little love for elected assemblies: Charles de Gaulle.

In Greece today, it could be argued that the decision about whether to accept the European Union’s bailout plan is no ordinary matter of economic policy. Perhaps it affects the country’s future as a sovereign nation so greatly that this indeed becomes an instance where sovereign power should revert directly to the people, making a referendum appropriate. Yet Greece has not had a referendum since the abolition of the monarchy in 1974. And consider what happened six years ago, when member states of the European Union were asked to consider a new European constitution—a question of sovereignty if ever there was one. France held a referendum—and voted no. So did the Netherlands. But in Greece, the parliament reserved the decision for itself, and gave overwhelming approval.

Given this context, Prime Minister Papandreou’s sudden enthusiasm for a referendum is not only undermining the world economy, but also Greece’s none-too-stable, 37-year-old democracy. It is political cynicism of a sad, but all too familiar sort—the kind that presents itself as idealism.

David A. Bell, a Contributing Editor to The New Republic, teaches European History at Princeton.