DR Congo runoff

The presidential election in the Democratic Republic of Congo is headed to a runoff in October. Incumbent Joseph Kabila won about 45% of the vote, and his closest challenger was nearly 25 percentage points behind.

Normally, I am skeptical about the value of rules requiring a runoff when the leading candidate is so far ahead of the closest challenger. In such situations there is usually little chance that a runoff could stop the plurality candidate. While Kabila will almost certainly win, the tense nature of the ceasefire between Congo’s various militias probably makes the jockeying for alliances between now and the runoff a valuable exercise (even though the possibility of a blow-up also remains present).

Additionally, such a large gap between the top two candidates often means that the second candidate has barely surpassed the third. In such a case, there is always the chance that the third candidate would be a more viable opponent to the leader than is the second candidate. This is an inherent flaw in two-round systems, especially in the context of political fragmentation. However, in this case, the political situation is so fragmented that the second candidate actually has quite a substantial lead over the third.

The votes shares of the top four candidates are: 44.8, 20.0, 13.0, 4.8. (The remaining 17.4% of votes is divided among twenty-nine other candidates, no one of whom has more than 3.5%! If even a few of those candidates also had legisaltive slates that won seats in their respective main districts, this will be one fragmented legislature!)

See Yebo Gogo for futher discussion, and a map that shows how regionally divided the parties are. The runoff campaign will be an opportunity for Kabila to try to build political support in the west, where he is much weaker.

UPDATE: At the propagation bench, Jonathan suggests that a victory by Kabila is a good deal less likely than I imply above. As I note in response to him, if Kabila loses, it just may be a record in the history of majority-runoff presidential elections!

0 thoughts on “DR Congo runoff

  1. Kabila will almost certainly win

    I wouldn’t be sure of that. The reason Kabila did so much better than Bemba is that he had a virtual lock on the eastern provinces, while Bemba split the west with Gizenga and several other also-rans. The third, fourth and fifth-place candidates all come from parts of the country where Kabila is despised, and (with the possible exception of Mobutu fils) they’re all likely to support Bemba in the runoff. The sixth and seventh-place candidates, who are from the east but anti-Kabila, are also tipped as part of the TSK (Anyone But Kabila) coalition in the second round. And then there’s the Tshisekedi factor – he’s nominally sitting out the election, but he’s expected to line up against Kabila and he has considerable clout across the country.

    Assuming that the endorsements in fact break this way, that would leave Kabila and Bemba at rough parity, with neither certain of a majority. This will set up a situation in which all the also-rans, down to the last-place candidate with 0.10 percent of the vote, will be able to participate in the bidding. Congolese politics over the next couple of months is going to be a seller’s market like nobody’s ever seen.

    If even a few of those candidates also had legisaltive slates that won seats in their respective main districts, this will be one fragmented legislature!

    You don’t know the half of it. There are – wait for it – 213 parties competing in the parliamentary election. According to article 115 of the electoral law, Kinshasa makes up 4 districts while each of the DRC’s cities and territories is a single district for a total of, I believe, 169 districts. In the single-member districts, election is by plurality, and in the multi-member districts it’s some kind of proportional representation (my technical French isn’t good enough to make sense of article 119).

    This arrangement favors small parties with local bases, and there are lots of those in the DRC. Given the number of parties that are based on personal and/or ethnic loyalties, I’d say that, oh, 50 to 75 of them have a fighting chance to get in. That should add a whole new dimension to the deal-making – after all, the incoming president will also have to form a broadly based government if he hopes to keep the country together.

  2. I’m from a state where all elections go to a run-off unless a candidate wins a majority. Honestly, I like it and it seems to work well — for many elections, parties forgo the primary elections (which tends to bug me) and, I believe, it often helps a more moderate voice emerge.

    Then again, I consider myself a moderate, so maybe that’s why I like it…

  3. If Kabila loses, I think it would be an all-time record. The all-time record comeback in a runoff for the second-place candidate was in Portugal in 1986. Center-right candidate Diogo Freitas do Amaral led the first round 46.3-25.4. Yet the first-round runner-up, Socialist Mario Soares, came from behind and won the runoff, 51.3-48.7. (The third candidate in the first round, supported by the Comminists and other left parties had 20.4%.)

  4. Fontaine, it is certainly true that majority runoff is more favorable to moderates than is plurality. Of the two who remain standing after the first round, the one that voters perceive as more moderate is far more likely to win than the one perceived as more extreme.

    Of course, therein lies the rub. Maybe the third candidate was the more moderate, yet she has been eliminated.

    I prefer that my runoffs be “instant,” though it does not guarantee the election of the more moderate candidate, either. It is somewhat more likely to do so, in that the “final two” can be different than after a first round of two.

    Some folks like Condorcet (simulates a series of one-on-one pairs), but it is not used anywhere for public offices and there are presumably good reasons for that. (This has been discussed in various threads here in the past.)

  5. If Kabila loses, I think it would be an all-time record.

    Kabila’s problem is that 45-50 percent of the vote may represent a maximum as well as a minimum. Half the country likes Kabila and half the country hates him; those who like him voted for him in the first round, and Bemba will pick up the great majority of the haters in the second. I’d say that, at this point, it’s too close to call.

    A big endorsement could still change that, though. There are rumors that Kabila might offer to make Gizenga prime minister (a post Gizenga last held a decade before Kabila was born) in exchange for an endorsement, which would shift the momentum Kabila’s way. The offer may not materialize, and Gizenga might not accept if it did, but the next government will pretty much have to be a Kabila-Bemba-Gizenga triumvirate anyway if the country is going to stay together, so why not start building it now?

    I prefer that my runoffs be “instant,” though it does not guarantee the election of the more moderate candidate, either.

    Instant runoffs are better in established democracies – the voting is done with in one day, there’s a chance for “everyone’s second choice” to be elected despite the lack of a primary support base, and the final arbiter of the election is voter preferences rather than back-room deals. In the DRC, however, a two-round system may be more suitable precisely because it encourages deal-making and doesn’t produce an instant winner. The cooling-off period between the first and second rounds gives losing candidates an incentive to sell their support rather than go into rebellion, and ensures that the eventual winner will be broadly and visibly endorsed. In an established democracy, that sort of bargaining is a bug, but in a country like the DRC which doesn’t yet have a democratic ethos and where civil war is a much more immediate worry than corruption, it’s a feature.

  6. Jonathan, your description of IRV as electing “everyone’s second choice” despite the lack of a primary support base would be more accurate for Condorcet or Approval Vote.

    IRV simulates a two-round runoff in one round of voting, and thus the candidates who survive into the late rounds of counting and vote transfers must have a strong primary support base.

    But I share your apparent skepticism about IRV in places with low literacy. And it is worth mentioning that the Supplementary Vote (an IRV-like ballot, but often with limited ranking opportunities, and which eliminates all but the top two before transferring votes) is NOT a viable alternative. SV has a very high probability of producing a Condorcet loser (the candidate who would have lost to any other in a one-on-one pairing), whereas IRV and regular runoffs make the election of a Condorcet loser almost impossible.

  7. IRV simulates a two-round runoff in one round of voting, and thus the candidates who survive into the late rounds of counting and vote transfers must have a strong primary support base.

    Point. IRV isn’t entirely a simulation of a two-round runoff, though. It’s possible for third or fourth-place candidates to win an IRV election on preferences, especially if their primary support is fairly close to that of the second-place candidate (example). I have a feeling, for instance, that Gizenga would have won an IRV election – he’s a lot more acceptable than Bemba to Kabila voters and vice versa.

    But I share your apparent skepticism about IRV in places with low literacy.

    I don’t think the problem is literacy, education or even level of development so much as it is democratic ethos. IRV has been used successfully in Fiji, where most indigenous people have only a primary education. For that matter, the literacy rate in the DRC approaches 70 percent and is especially high among youth, and IRV there would be a disaster. I’d argue that the key lies in whether elections are the accepted method of transferring power or whether losers typically respond by secession and/or rebellion. If the latter, then a system that encourages deal-making by losing candidates is preferable to one that doesn’t.

    With that said, I’ll be interested to see how PNG does with the system in 2007.

  8. I have a feeling, for instance, that Gizenga would have won an IRV election

    Gack. I’m confusing systems again. What I think I meant to say is that Gizenga would be the Condorcet winner. Given that the also-rans were beginning to divide into Kabila and Bemba camps even before July 30, I expect that most of their second preferences would have gone to Kabila or Bemba, and that one of the two leaders would thus have won an IRV vote.

  9. Jonathan added: “It’s possible for third or fourth-place candidates to win an IRV election on preferences, especially if their primary support is fairly close to that of the second-place candidate”

    Yes, of course. I did not mean by “simulate” that IRV would result in the same two candidates being the last two left standing after all other votes were transferred. Of course not, and that’s one of the reasons it is more likely than regular two-round system or SV to produce the Condorcet winner. If the candidate in third place on the initial count gets more transfers from lower-ranked candidates, then he might pass the initially second candidate, leading the latter to be eliminated.

    My remark was a reaction to the phrase “lack of a primary support”–apparently an overreaction!

    As Jonathan says, if the third or fourth candidate is close to the second in the first count, then one of these might win. That’s a good thing, as one of the drawbacks of two-round systems (as I have noted here with respect to Peru and other cases) is that sometimes a candidate who narrowly missed the runoff would have been the stronger challenger to the leading candidate.

  10. This is late as a result of the meltdown, but the Australian parliamentary library has compiled a database of IRV elections for the house of representatives where the leading candidate on primary votes was displaced by another candidate after transfer of preferences.

  11. On the legislative elections, Jonathan Edelstein wrote: “in the multi-member districts it’s some kind of proportional representation (my technical French isn’t good enough to make sense of article 119).”

    Let me help: it’s list-PR with simple quota and largest remainder (“la règle du plus fort reste”). In small constituencies where votes are spread over many lists, this can give strange results.

    Within the lists, seats are distributed purely on the basis of personal votes – one could only vote for a specific candidate on a list (“listes ouvertes à une seule voix préférentielle”)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.