The Bolivian political crisis set off by credible (if unconfirmed) allegations of fraud in the October 20 presidential elections took two dramatic turns this week. On Wednesday, a major escalation in clashes between pro- and anti-government demonstrators in downtown Cochabamba and near Huayculi in Cochabamba caused scores of serious injuries and the death of a twenty-year old student, Limbert Guzmán, who had been protesting against electoral fraud. The anti-government side engaged in less-lethal outrages upon government-aligned actors: subjecting the MAS-IPSP-affiliated Mayor of Vinto to march to the site of the clashes and pelting her with red paint, and last night burning out the Cochabamba offices of the Chapare coca grower’s union and MAS-IPSP political party.

But the pivotal event of yesterday is the decision of members of the national police to declare themselves in mutiny in solidarity with the electoral fraud protests. (Police mutinies are a periodic occurrence in Bolivia, and the term does not connote a necessarily violent uprising, but rather a collective refusal to follow orders.) The mutiny began at Cochabamba’s Police Operations Tactical Unit (Wikipedia), a specialized anti-riot force, and quickly spread to two other units in the city, and to police units and/or commanders in five other cities since then. Mutinied police officers have occupied their own barracks, raised Bolivian flags and sometimes anti-fraud banners, and effectively removed themselves as an option for President Evo Morales to control mushrooming protests. The ongoing protest mobilization has gathered hundreds or thousands of protesters around police barracks supporting existing mutinies and encouraging further ones. Meanwhile, the Morales government has described the police mutiny as the unmistakable sign of a coup d’état.

Police mutinies in their bottom-up form have accompanied Bolivian protest before. They can force governments to negotiate and avoid bloodshed. Or, in the worst case (Feb 2003), there can be bloodshed among state security forces. The next critical question is the stance of the military. Early signs are neither the military leadership nor Minister of Government appear to want the police and military to be deployed against one another.

Bolivia has had a lot of off-schedule changes of power, but since 1982 the have always been ratified by a vote at the ballot box. This was true in mid-1985, when general strikes and business lockouts forced Hernán Siles Zuazo to call early elections; in October 2003, when the Gas War uprising, supported by a series of hunger strikes led Gonzalo Sánchez de Lozada to resign; in June 2005 when further protests pressed Carlos Mesa to resign, and two legislators to forego their right to presidential succession, prompting December 2005 elections; and during the 2008 political crisis, when dueling protest mobilizations were ultimately resolved in favor of the Morales–grassroots left coalition, with the outcome ratified by a recall referendum and a constitutional referendum. In April 2000, a police strike influenced the government decision to capitulate to the Water War protests in Cochabamba and de-escalate its repression of rural protests in the Altiplano. At the final moments, the military high command pressed for a constitutional exit to the crises in 2003 and 2005.

In all of these cases, the military signaled limits to further state repression, stayed out of the presidential chair, and did not substitute its choice of leaders for one determined at the ballot box. Arguably, this is why none of these events are remembered as coups d’ètat. In the current context, doing so will require holding some kind of vote under terms that are acceptable to a broad swath of the Bolivian public and that acknowledges the rights of both Morales and Mesa voters. If and only if the police, military, Morales government and electoral opposition can agree to such an exit, democracy can be preserved.