The Vote That Cannot Write the Peace

That is not a defective ballot. It is the limited instrument Colombia has chosen for a very large decision. Negotiation produces a package; ratification asks the public to accept or reject it. When the answer is rejection, the political system receives a binding instruction without receiving a replacement agreement. The counting is considerably easier than the interpretation.

On October 2, 2016, Colombian voters rejected the peace accord between President Juan Manuel Santos’s government and the Revolutionary Armed Forces of Colombia, or FARC. Reuters, reporting through El Financiero, said that with 99 percent of votes counted, the electoral authority reported 50.24 percent for “No” and 49.75 percent for “Yes.” Those are the report’s provisional figures, not a final certified tally.

Four years of negotiations in Cuba had reached a public veto. The conflict the agreement sought to end had lasted more than half a century. Neither duration cancels the other: years spent negotiating cannot require public approval, and public rejection cannot make the costs of continued conflict disappear.

What the Voter Could Actually Decide

The scope of the choice matters because the language surrounding it was already larger than the legal question.

An Associated Press report published that morning framed the alternatives as a path toward peace or a continuation toward war. That conveyed the stakes, but it also compressed a distinction essential to understanding the result. Rejecting these terms is not necessarily a preference for fighting. Nor does describing rejection as a demand for better terms establish that better terms can be secured.

The Constitutional Court had addressed precisely this boundary before the vote. In judgment C-379/16, it distinguished the special plebiscite from a referendum that directly inserts a legal text into the legal order. The public was being asked to approve or reject a presidential political decision, one whose implementation would require subsequent legal action.

The Court also defended treating the agreement as a whole. Its explanation was not that every provision was beyond dispute. It was that the settlement contained interconnected matters that could not simply be separated for approval without disarticulating the agreement and requiring renegotiation. A voter who disagreed with part of it therefore had to weigh that objection against the whole.

Return to that voter. The institution allows a consequential refusal, but not an amendment. Whatever combination of approval, distrust, hope or objection precedes the mark on the ballot, the counting process records one answer. This is useful for authorization. It is a poor instrument for diagnosing what would make authorization possible next time.

That limitation should restrain both camps. Supporters cannot infer that every opponent has rejected peace. Opponents cannot infer that every “No” endorses the same alternative. The result settles whether this proposal won the required support. It does not settle what coalition could support another.

The Agreement Was More Than a Ceasefire

The breadth of the package helps explain why its rejection cannot be translated into one policy instruction.

The morning AP account described a 297-page agreement covering disarmament and verification, the FARC’s entry into political life, land distribution, illicit crops, political participation and redress for victims. It also described arrangements involving amnesty and the pursuit of responsibility for wartime harm. These were connected features of a proposed transition, not accomplishments already delivered by signing a document.

Each raises a different practical question. A person seeking redress needs to know how responsibility will be established and what remedy can be obtained. Someone asked to surrender a weapon needs credible arrangements for the transition out of armed life. A community promised changes concerning land or illicit cultivation needs to know what institutions and resources will make those promises usable.

These are tests for evaluating the agreement, not claims that any one concern explains the national result. The sources considered here do not establish a single motive shared by the “No” electorate. A close national count cannot supply that missing evidence.

Yet the institutional difficulty is visible without guessing at voters’ minds. A negotiated settlement joins commitments that may depend on one another. Removing an unpopular provision can change what another party is willing to accept. Leaving it untouched can preserve the bargain while preventing it from acquiring public authorization. The document must survive both encounters.

Calling the public unreasonable does not solve that problem. Calling the negotiators traitors would not solve it either. Both gestures replace the work of constructing an acceptable settlement with the more familiar work of identifying someone to blame. Blame is administratively convenient. It requires no implementation budget.

A Binding Refusal With Defined Limits

The result cannot be treated as advisory merely because it frustrates an important objective.

In paragraphs 114–117 of C-379/16, the Court explained the consequences of rejection. An adverse vote would prevent the president from implementing that specific agreement. The Court rejected an interpretation under which only a favorable verdict would have substantive consequences. Such a reading would reduce public participation to a device for validating government policy.

This is the hard edge of consultation: if “No” cannot alter the course, the public has been invited to applaud, not decide.

But the Court drew another boundary just as firmly. The plebiscite did not put the constitutional right and duty to peace, or the government’s constitutional powers to pursue it, up for removal. Those remained intact. The Court expressly contemplated a new decision with different conditions, arising from renegotiation or a new agreement, being put to the people under the special plebiscite framework.

Its operative ruling also confined the binding effect to the president rather than treating the vote as an automatic command rewriting the responsibilities of every branch of government. These distinctions are not decorative qualifications. They describe what remains possible after rejection and what cannot simply proceed unchanged.

As of this vote, a route contemplated by the Court is not a negotiated outcome. There is no warrant for assuming another agreement will emerge, that it will command approval, or that the parties will agree on what must change. The legal room to continue working is real; the political success of that work is uncertain.

For the voter, this produces a demanding but coherent position: the refusal must have force, and the state must continue pursuing peace. Neither obligation can be used to erase the other. The executive does not get to treat the result as a clerical inconvenience. Its opponents do not acquire, through the same result, a constitutional instruction to abandon the pursuit of a settlement.

Reaching the Ballot Is Part of the System

Before any vote can express a preference, someone has to be able to cast it.

An AP report on the closing of the polls described flooding that hampered the installation of polling stations in parts of the country. Senator Armando Benedetti and several governors sought an extension of voting hours amid rain associated with Hurricane Matthew. The National Electoral Council declined to extend the day. A magistrate said the council had not found sufficient legal grounds for an extension.

This reporting establishes a problem of access and an institutional response. It does not establish how many votes were prevented, how those people would have voted or whether a longer day would have changed the result. Weather must not become a convenient substitute for evidence about the margin.

Still, access is not incidental to legitimacy. A consistent closing time and a usable opportunity to vote are both institutional goods; under disruption, they can pull in different directions. Reporting the rule alone loses the person trying to reach the polling station. Reporting the obstacle alone does not tell us what lawful remedies were available.

The same concern begins earlier than polling day. In its ruling, the Constitutional Court required publication and dissemination of the agreement to account for people with disabilities and communities that do not communicate in Spanish. Its reasoning emphasized remote communities and groups heavily affected by the conflict. It recognized that making a document available is not identical to making its contents accessible.

That requirement is not proof that dissemination succeeded or failed everywhere. It supplies a standard against which the process should be examined: could people understand the decision before being asked to authorize it? A public choice is thinner when the institution counts the answer but neglects the conditions needed to form and express it.

The Work After the Count

A coherent response must now do more than announce respect for democracy or devotion to peace. Those declarations are compatible with almost any subsequent maneuver.

First, the rejection must remain attached to the decision actually rejected. The Court’s account makes clear why proceeding with the same agreement as though nothing happened would empty the consultation of meaning. Equally, expanding the verdict into a rejection of all negotiation would attach to the ballot an instruction it did not contain.

Second, political argument must become specific enough to support revision. Which provisions are contested? What changes would address those objections? What reciprocal commitments would those changes disturb? These questions are slower than claiming to speak for “the people,” which is one reason that phrase is so popular after close votes.

Third, people carrying the consequences of delay and conflict must remain inside the assessment. Their claims cannot be reduced to material for either campaign’s explanation of the result. Nor should a negotiated peace purchase its appearance of unity by making objections, victims’ demands or practical implementation problems inaudible.

The test is not whether everyone agrees. It is whether the next process can receive those objections, distinguish among them and show what changed because they were heard. Feedback matters when it can alter a proposal—not when it is collected and filed beneath the announcement of consensus.

Colombia’s vote has demonstrated the power to withhold authorization. It has not supplied a common plan for what follows. The next danger is that leaders will mistake possession of a result for possession of that plan. A durable peace will require them to do the harder thing: keep the refusal binding without allowing it to become the last available political act.