In presidential systems, the rare “counterhoneymoon” electoral cycle–legislative elections shortly before presidential–can have the effect of revealing the strengths of various parties and their prospective presidential candidates before the final phase of the presidential campaign itself. I have long thought that a counterhoneymoon cycle was normatively desirable, especially for multiparty systems, and have been surprised at how rare it is.

The effect is on display now in Colombia. For many months leading up to the 14 March, 2010, legislative elections and 30 May presidential first round, Sergio Fajardo, former mayor of Medellin, was doing well in the opinion polls that asked about hypothetical match-ups without incumbent Alvaro Uribe. (Uribe was recently denied a chance to run for a third term by a court ruling.)

However, the outcome of the legislative elections has changed the dynamic. As noted in El Tiempo:

Aunque llegó a liderar las encuestas sobre intención de voto para Presidente de la República, hace apenas tres meses, Fajardo se desplomó tras su fracaso en las elecciones legislativas, pues ni siquiera uno de los suyos llegó al Capitolio.

[my rough translation:] Although he was leading the polls of voting intention for president only three months ago, Fajardo fell flat after his failure in the legislative elections, when not even one of his candidates made it to the Capitol.

Fajardo has fallen so far that now he is in talks with the nominee of the new Green Party, former Bogota mayor Antanas Mockus, about preparing a unity ticket for the presidential elections. (This is the main theme of the above-referenced article, and also the theme of Steven’s earlier planting.) If in coalition with other non-uribista forces, Mockus might even have a chance at finishing second in the first round of the presidential election, behind the successor to Uribe in the “La U” party, former Defense Minister Juan Manuel Santos.

Mockus decisively defeated two other former mayors of Bogota in a presidential primary held concurrent with the legislative elections. The other main presidential candidate is that of the Conservative Party, Noemi Sanin, who narrowly edged out a more pro-Uribe contender, Andrés Felipe Arias, in that party’s primary.

While there is believed to be no chance that Fajardo would become Mockus’s running mate, there is speculation that Fajardo’s already-registered running mate, Julio Londoño, could now run with Mockus instead.

What is the broader significance of all this? I often write about “presidentialized parties,” by which I mean parties that become dependent on their leaders who run for president. In fact, I have co-written a book about such parties, and their fundamental difference from “parliamentarized parties,” in which the agency relationship running from parties to their executive candidates is more readily maintained, even after an election results in the candidate becoming the national executive (precisely the moment in which the agency relationship most breaks down, or even reverses, in presidential systems).

Few parties are more “presidentalized,” at least prior to their actually having elected a president, than are parties that are formed for the sole purpose of being some leader’s campaign vehicle. Compromiso Ciudadano por Colombia, formed by Fajardo, for example. (Fajardo is registered as an independent, but that is a mere technicality for present purposes.)

Yet the counterhoneymoon electoral cycle has the effect of requiring such a party’s first test to be a legislative election, rather than the presidential contest. Fajardo’s party failed that test badly, and now he is in negotiations with the leader of a party that actually performed respectably in the legislative elections, notwithstanding that it, too, was just formed in time for these 2010 elections. (The Greens won 5 of the 100 Senate seats, for example.)

Other recent examples of counterhoneymoon effects: the 2009 elections in Indonesia and El Salvador.