The Dangers of Giving the Common Man a Say

“There go the people. I must follow them, for I am their leader,” the French politician Alexandre-Auguste Ledru-Rollin once said memorably and, perhaps, apocryphally. Even if the quote isn’t his, there’s a reason why so many have come to believe otherwise. It represents a leadership style much in vogue during his time.

Ledru-Rollin came to prominence during the Revolutions of 1848. Europe was in turmoil: Countries across the continent were witnessing revolts against monarchies that had in some cases been in power for centuries, from Sicily to Hungary. There was an explosion of popular participation in politics, and politicians like Ledru-Rollin acquired power when the old rules of being in charge seemed to no longer apply. So rather than taking the lead and making decisions, they turned to plebiscites. In the confusing decades after 1848, referendums popped up everywhere; in France, they were held to instate Napoleon’s nephew as emperor; in Italy, to unite the country.  There was a lull before World War I, but the years after the war saw another wave of referendums, especially on contested borders between Germany and its neighbors Denmark and Poland.

In general, the use of referendums is characteristic of times of upheaval — periods when the elites are uncertain of their support. Referendums were held in the wake of the French Revolution, the Russian Revolution, World War II, and the fall of communism. They followed the re-establishment of democracies in Greece in 1974 and Spain in 1976.

The problem is that, in times of upheaval, these plebiscites are as likely to backfire as to help politicians get their way, or consolidate support. That was true for Napoleon I, Charles de Gaulle, David Cameron — and now it is true for President Juan Manuel Santos of Colombia and Victor Orban of Hungary. In the Colombian referendum,  voters were asked to endorse a peace plan between the Revolutionary Armed Forces of Colombia (FARC) and the government, and declined. In the Hungarian case, the government asked voters to reject the European Union’s plan for redistributing refugees across the EU; they did, but failed to turn out in sufficient numbers to render the vote valid.

In times of stability, political leaders know their audience and their followers. From the 1970s to the 1990s, political elites, broadly speaking, felt confident about their hold on power. They had relatively united parties behind them. The world was neatly divided into left and right. They were — again, broadly speaking — trusted to make decisions.

And so they did — including decisions very similar to the one Santos handed over to voters this month. There was no referendum on the Esquipulas Peace Agreement in 1987, which ended years of fighting in Nicaragua. Nor was there a referendum on the Anglo-Irish Agreement in 1986 — a major step in the Northern Ireland peace process (though there would be plebiscites held to approve the Good Friday Agreement 12 years later). Overall, it seems likely that, had the Colombian government managed to end its five-decade war with the FARC a few decades earlier, it would not have sought the approval of voters to do so.

Now, however, the world and its politics are in flux. No, we are not quite experiencing a time like the aftermath of the 1848 revolutions, but long-standing left-right fault lines taken for granted for decades no longer seem to apply. Just consider the recent study by the Swedish academic Sören Holmberg that showed voters agree with only 70 percent of “their” party’s platforms. As a result, insecure politicians are once again turning to referendums as tools. Even the Scottish National Party — a party built around Scottish independence — held a referendum on the question in 2014, despite having gained a majority three years earlier (and, indeed, when asked, a majority of Scots voted against independence). If the SNP no longer views its majority as a mandate to pursue independence, perhaps we should not be surprised that Santos felt it was not politically feasible to make peace with the FARC without consulting voters first.

But from the perspective of the politicians who organize them, there’s a problem with referendums: They don’t control the outcome. Sometimes ruling parties can lose unexpectedly, which happened in Colombia, where most expected the peace deal to sail to an easy victory. Or, in the case of Hungary’s Orban, who also held a referendum that failed to draw sufficient turnout to produce a binding result, they can be won — but not in the way a government might have hoped.

But the fact that referendums are not predetermined doesn’t mean they’re not predictable. Although politics is not an exact science, recurrent patterns are instructive. The main factor is that referendums are often votes of confidence or no confidence in the government. One thing we know about referendums is that governments that have been in power for a long time tend to lose them more frequently: On average, governments lose 1.5 percent of support for its positions on referendum votes for every year the government has been in office. In other words, voters lose faith in politicians who aren’t fresh off their mandate. Santos — a two-term president halfway through his second term — may have been in office for too long to feel confident of victory.

Another consistent factor is that referendums are generally lost if there is a politically credible politician standing on the side of the vote that opposes the government. In France, in the referendum on the European Constitution in 2005, the declaration of former Prime Minister Laurent Fabius that he would vote “no” made it impossible for President Jacques Chirac and the “yes” side to depict opponents of the treaty as extremists. The same may have been true in Colombia: It was difficult to portray opponents of the peace agreement as extremists after the former, but still respected, President Álvaro Uribe declared his opposition. And so, in fact, while pollsters may have expected a victory in this weekend’s vote in Colombia, political science actually pointed toward a likely loss.

The result of the Colombian referendum might not have been foreseen by pollsters. From a political science perspective, the result in Colombia was not a surprise. Given Santos’s time in office, and Uribe’s support, it was almost a foregone conclusion.

Now the political elites have to go back to the drawing board — unless they decide to hold another referendum essentially posing the same question. Historically, they would have good reason for optimism — another general tendency is that the referendums held the second time around often get approved. That is the lesson from Denmark, Kenya, and Ireland, where voters vetoed policy proposals only to endorse new proposals that had removed controversial elements.

Of course, there is another possibility. Maybe politicians in Hungary and Colombia should be braver. Maybe they should present bolder policy platforms during their next general election campaigns and accept the verdict of the voters on those platforms as a mandate to enact them. And maybe, just maybe, they should think twice before holding referendums.

Photo credit: Oli Scarff/Getty Images