There’s three weeks before the deadline for parties & candidates to register for the December general elections in Bolivia. So far, the enduring calls by opposition candidates to join together in a “unity front” against Evo & MAS failed. Currently, a dozen opposition candidates are preparing to run against Evo.

Most of these figures don’t have much of a chance. The best candidates are so far polling only close to 10% while Evo is still the clear frontrunner with a comfortable 40-45%. Because of Bolivia’s new system for electing presidents, Evo can win reelection with only 40% of the vote if he beats the second-place candidate by at least 10 points.

At least two of the candidates, Tuto Quiroga & Doria Medina, have faced off against Evo before. This makes for an interesting repeat of the December 2005 elections. Looking at old polls is interesting. Between February 2004–July 2005, Tuto regularly led in the polls, w/ Evo & Doria Medina close together (see image). By September 2005, Evo had surged ahead to become the frontrunner, w/ Tuto & Doria Medina virtually tied for second (see image). By October 2005, Evo continued to gain ground, but Tuto & Doria Medina were still close (see image). In the weeks before the election, Evo was clearly ahead, w/ Tuto & Doria Medina at about where they ended up (see image). In the final poll days before the December 2005 election, Evo continued to gain, while Tuto & Doria Medina continued to drop (see image).

What’s interesting about this last poll, however, is that it grossly undercounted Evo’s support (by about 20 points!) even though the final results for Tuto & Doria Medina were on par w/ the polling numbers. One explanation is sample bias: Bolivian polls look almost exclusively at cities. While these account for nearly 2/3 of Bolivian voters, their preferences are almost certainly structured differently. Another explanation is to look at the “don’t know” respondents in polls. If they’re taken out of the equation, the numbers change. Such respondents may either not show up to vote or vote blank/null or decide at the last minute or be hiding their preference from pollsters. Either of these (or a combination of them) could produce a systematic bias favoring a candidate.

Bottom line: Bolivian polls are often wildly inaccurate. The numbers rarely matter as much as the trajectory. In the December 2005 polls, it mattered little what Evo’s actual numbers were. What mattered was that they consistently went up (and his opponents’ went down) after September 2005.

As the candidates go into full swing in Bolivia, it will be interesting to see what happens in the polls. This time, Evo is starting w/ the advantage of incumbency (which can also become a curse) & higher initial poll numbers. A June poll gave Evo 44%, up from 41% in April. But those numbers were 10 points below his approval ratings for the same month (see my earlier post about this). The latest poll gives Evo 45%, followed by a fragmented opposition field (see image). This latest poll also asked voters about a hypothetical “unity candidate” for the opposition. The responses to such a questions till gave Evo a comfortable 48-33 lead (enough to secure reelection, if those were election results). But it’s interesting that Evo’s numbers only increased 3 points relative to the full field. This suggests that while most (urban) respondents support Evo, those who don’t would oppose him irrespective of their next option. That’s a highly polarized electorate.

This time, in addition to Tuto & Doria Medina, two other “strong” candidates will challenge Evo: Victor Hugo Cardenas & Manfred Reyes Villa. A respected indigenous intellectual & former vice president, Cardenas has significant support among La Paz middle class. A four-time mayor of Cochabamba (Bolivia’s third largest metropolitan area) & populist machine politician, Manfred has significant support from his political machine, as well as sectors of the traditional/populist right.

So far, it seems unlikely that the opposition will form a unity front. Each of the chief candidates clamors for one. But I’m certain the sticking point in the negotiations is that each wants to be the presidential candidate. Additionally, the opposition has spent the last few years in self-imposed disarray, making little effort to put together a political platform. The most “successful” opposition since 2005 has come from the Media Luna. But demands for regional autonomy (now met in the new constitution) are not enough for a national-level political program. It’s also no surprise that none of the incumbent prefects will run for president; they’ll instead seek reelection as governors (the new term for prefects in autonomous departments).

Of course, it may matter little what candidates do. Voters may, on their own, decide to flock to one of the opposition candidates. This seems to have happened in 2005. As the election neared, Doria Medina’s poll numbers declined rapidly. The final results suggest that most of Doria Medina’s voters shifted to Tuto. Such shifting could happen again, especially if polls are frequent & widely diffused. But that kind of electoral coordination is difficult to achieve, and will never work as well as simply presenting voters w/ a binomial choice (Evo v. “unity candidate”). Right now, the opposition is plagued by “too many chiefs, not enough Indians” syndrome. That’s not the case w/ MAS. While it’s not a strongly institutionalized party, there’s clearly a strong leadership in MAS: Evo Morales.

Either way, it matters little whether the opposition can forge a unity slate. Evo will almost certainly win reelection. But the obsession w/ the presidency means that opposition figures are ignoring the importance of the legislature. Here, strong opposition parties (rather than one-shot electoral vehicles) are essential. What the opposition should focus on is building new (or revitalizing existing) parties that can articulate a political program. Even in defeat, such a party can work in the legislature effectively (Podemos, despite controlling the Senate for three years has managed to do little). The opposition should take a long view, working to build a party system that can secure a democratic Bolivia into the foreseeable future by offering an institutional “check” on the ruling party. Right now, the worse thing that could happen to Bolivia is for an opposition candidate to actually win the presidency in an upset. That would bring us right back to 2003. The combination of a weak president, a fractured party system, and strong social discontent in the face of mounting international economic problems would spell disaster.