Do divided oppositions make incumbents win ?

December 29, 2007

In Africa, that’s it.

You’ve heard it before: too many candidates, the opposition split its vote, blah blah.. It’s right after rigging, media control and pure exclusion of the opposition, one of the most popular ways to explain why so many leaders keep winning elections. Here I’ll assume there’s no rigging (and no candidate exclusion) and I’ll concentrate on presidential elections (who cares about parliaments after all ?).

Four countries on the continent don’t have a president, three of those are monarchies (Morrocco, Lesotho and Swaziland) and one is Libya. Also, Eritrea is an extra-legal dictatorship and Somalia has no government.
In four countries, the president is elected by the parliament (or the congress): in Ethiopia, where the president is ceremonial, in Burundi which is a transition period, in Botswana and in South Africa that I’ll discuss later.

In order for the “split opposition” theory to work, it has to be possible to win even if more than half of the voters didn’t vote for you. That implies it cannot happen in countries with a run-off/two-rounds system. After all, no matter how divided the opposition is, they have the option to regroup during the second round. It has happened in 2000 Senegal election where challenger Abdoulaye Wade won 58% of the votes after getting 31% in the first round while Abdou Diouf had the same score 41% in both rounds or in the 1993 Niger elections where the incumbent was by far the single biggest party (both in presidential and legislative elections) before the opposition formed a large majority coalition against it.
Of course, that doesn’t always happen. Sometimes the incumbent does an even bigger score during the second round like in DRC 2006 or Niger 2004 mostly because some of the other candidates and factions decide to endorse him. I’m not sure that should count as a case of divided opposition. If anything, the incumbent’s vote could be divided too if some of the minor candidates have platforms close to him. For instance, in the congolese case the third man, Antoine Gizenga – Lumumba’s deputy prime minister in the 60’s – , probably had more in common with Kabila – the son of a lumumbist revolutionary – than with Bemba – the son of a Mobutu apparatchik – (the fact that Mobutu’s own son joined Kabila’s coalition support the alternative view).
It’s also interesting to note that in the most authoritarian of those countries, the incumbent wins after the first round and by staline-era margin. I fail to see how 5 candidates who getting 5% each are any worse than one who gets 25%, the bad guy still got 75%.
And the prevalence of TRS ? Well, that’s more than half of Africa including all of Francophone Africa (except Gabon who got rid of the second round in 2005 for “fiscal” reasons, not that Bongo ever got close to facing a run-off), Ghana since 1976 and the war-recovering new democracies of Liberia, Guinea-Bissau or Sierra Leone (where a candidate has to win 55% of the vote to avoid a run-off).

Similarly, South Africa uses a proportional system for its parliament. If ANC ever gets less than half of the vote, nothing stops the rest of the parties to form a majority after the election.

But then you have the First Past The Post countries where it is possible for someone to get elected even if the majority voted against him, as long as he has a plurality. There’s thirteen of the countries, aforementioned Gabon and elections-delaying Angola included.
In most past cases, the winner did have a majority of the votes which makes the number of opposition candidates arithmetically irrelevant. In Seychelles, for instance, former sole legal party SPPF has won every single election with a clear majority, no matter how many parties/candidates ran against it. Tanzania, Botswana, Mozambique, Gambia or Uganda are in the same case. Zimbabwe would have been too had the ruling ZANU not needed additional appointed seats to have a majority after the 2000 parliamentary elections.
That leaves us with exactly four countries where an united opposition could have made or did make a difference in the final outcome: Nigeria, Malawi, Zambia and Kenya.

In the 1979 Nigeria election, there was no incumbent as it was a return to civilian rule election. However, Shehu Shagari was the military’s favorite and won the election with just 34% of the votes beating Obafemi Awolowo’s 29% and Nnamdi Azikiwe’s 17%. Four years later, Shagari was re-elected with 47%, while Awolow had 30% and Azikiwe 14%. An Awolowo-Azikiwe common platform could have won the election and wasn’t a far fetched idea as they had been part of the center-left United Progressive Grand Alliance (Azikiwe leading NCNC and Awolowo leading Action Group) that got defeated in the 1964 parliamentary election. However, that is assuming that Shagari’s vote was maximized both times and it wasn’t. As a matter of fact, both Awolowo and Azikiwe had the luck of being the sole candidates of their respective ethnic groups (Yoruba and Igbo) while Shagari had to face other northern/muslim/hausa candidates. And the two other northern candidates in 1979 got 20% of the vote and the three other northern candidates in 1983 got 10% of the vote. Given the ethnic vote patterns so prevalent in Nigeria (the 1959 results look like a census with minor alliance switches) at the time, it is fair to assume that it’s Shagari vote that got divided, not the opposition’s. Quite oddly, the cancelled 1993 election, which came after an ambitious political engineering process that involved the creation of two state funded and ideologically delimited national parties, managed to at least partially break the ethnic voting reflexes. While the winner, MKO Abiola, a Yoruba, did have large margins in southern states, he also won or narrowly lost many northern states. It seemed like many northerns realized having a center-left president was more important than having a northern one (the fact that Abiola was muslim and that both candidates had running-mates from the other side of the country probably had an impact too). Later elections produced interesting outcomes like Olesegun Obasandjo sweeping nationally while loosing his home state or open negotiations and campaigns about the ethnic origin of his successor. Basically, PDP, the ruling party, is the sort of large, flexible, all-encompassing grand political alliance that are really hard to challenge, especially if the challengers are frindge parties/candidates (like quasi-fascist Buhari) or rogues who can’t conceile their personnal ambitions (like former vice Atiku).

In Zambia, the opposition didn’t take any chances during the first multiparty election, it went to the polls united in the Movement for Multiparty Democracy and its candidate won 75% of the vote and ousted independance hero Kaunda. He would get reelected with a similar score 5 years later but former single-party UNIP boycotted the election and the other candidates were, i think, former MMD members. However in later elections, MMD’s candidate won by far less convincing margins. Less than 30% and just 2% more than the runner-up in 2001 and 42% in 2006 (but that’s incumbent bonus). I’m not totally up to date on the details of Zambia’s political situation but I do suspect that the opposition vote was as split as the ruling party vote. After all, 2006 runner-up, Michael Sata left MMD after not being picked as the presidential candidate, on top of having been close to Kaunda in the 1980’s.

In Malawi’s first multi-party election, an opposition divided between 2 candidates wasn’t enough to save incumbent Hastings Banda as Bakili Muluzi (UDF) got ahead with 47% of the vote. How am I sure the opposition was divided ? Well, Banda’s score (33%) in 1994 and the number of people who voted against multiparty rule in the 1993 referendum (35%) are virtually the same. Then Muluzi got re-elected with a 52% majority in an election where the former ruling party (MCP) and the other half the opposition (AFORD) had a common candidate. That move seemed to have badly hurt AFORD as in the 2004 election, they had been replaced by another “third way” coalition that gathered a quarter of the vote tying MCP but behind UDF’s candidate 35%. The real odd moment was Bingu wa Mutharika’s defection from UDF less than a year after the election. That political situation with a president with no allies in a parliament divided between three parties has a lot to do with some policy decisions in Malawi.

Finally, there is the only case where an incumbent got re-elected because of a split opposition. Daniel Arap Moi managed to get won with just 36% of the vote in 1992 just ahead of three opposition candidates who would have had 63% of the vote had they not runned separately. It’s also telling that two of them represented factions of a party that had split just months before the election. In 1997, Arap Moi did a better score, 40%, but once again could have been beated had Kibaki (31%) or Odinga (10%) dropped out of the race. In 2002, having learned its lesson, most of opposition formed a large coalition and its sole candidate, Kibaki won a 62% landslide. Though Arap Moi couldn’t run, his appointed successor had plenty of name recognition as he was independance hero Jomo Kenyatta’s son. However, the opposition honeymoon was short-lived. Government disagrements and a constitutional referendum split it and last thursday Kibaki and Odinga once again ran against each other. This time, however, a come-back of the big evil KANU is very unlikely and other issues make the election complicated.

So, no, assuming there’s no rigging, split votes rarely change the outcome.

But to be fair, there are other ways in which incumbents can benefit from a divided opposition. First of all, closer elections are harder to rig, so are Run-offs. Also a divided opposition looks unprincipled. The failure to put personnal ambitions apart for the greater good tells voters that no candidate has a monopoly on integrity and that makes voting for the devil you know or along ethnic lines or not voting at all very tempting. Also incumbents have the state apparatus working for them which gives them more name recognition, an extensive political network, the state ressources and media exposure. An opposition coalition could have all of those but a bunch of lone rangers only increase the noise-to-signal ratio in political discussions.
But all in all, united or divided, oppositions (and former ruling parties) have real staying power and chances only when they manage to make sense politically. Yes, that means ideology. In Cape Verde, for instance, former opposition coalition MpD managed to stay relevant after changes in leadership and defeats because it became a center-right party that represents more than an alternative to the bad guys. And former sole party PAICV managed to come back to power, in the tightest presidential election ever, because it’s clearly a center-left party. In Ghana or Senegal too, the political field is quite clear (and I’d argue that it is in Ivory Coast one of these days) and elections meaningful. And in Seychelles, Botswana or Mauritius the lack of political alternance is really a reward to incumbents, as there too, parties do stand for something.
The real problem then is facing politically vague, establishment, all-encompassing, “unity” parties like Nigeria’s PDP, Tanzania’s TANU, South Africa’s ANC, Namibia’s SWAPO, Rwanda’s FPR or Uganda’s NRM. Sort of like Mexico’s PRI (before Vox), India’s Congress, the Peronistas, or Malaysia’s and Singapore’s ruling coalition, they are virtually the only national parties with an extremely wide (geographical, social) reach, an historical role and – because they incorporate politically diverse factions – an incredible policy flexibility (how can you attack Mbeki or Museveni from the right when they’re viewed as pro-business by the business community and by some of their disapproving supporters ?). What serves as an opposition to those are fringe parties (Inkhata ? Azania People’s Organization? Freedom Front Plus? a coalition of those?) usually perceived as un-fitted to rule and those congress parties have to mess up real bad or experience a meaningful split in their coalition to ever loose. And more often than not, they do come back. So opposition leaders and supporters, tighten up your platform before making deals and day-dreaming about unity.

PS: I just realized that political systems in Botswana and Ethiopia strongly favour the ruling parties. Lower Houses are elected in single member constituencies using plurality. BDP’s share of seats in the parliament is consistently much larger than its share of the votes (scroll down). And in Ethiopia, I would bet that the opposition vote was large but concentrated in the (under-represented) cities where it has won all of its seats.


One Response to “Do divided oppositions make incumbents win ?”

  1. R. Sherman Says:

    Sorry for shotgun commenting but I just discovered your weblog and am enjoying myself.

    I cannot speak to Africa, but I note the problem you describe is not unique and is likewise not really that much of a worry here.
    Bill Clinton won his first term (1992) with a plurality of the popular vote as against Bush the Elder and Ross Perot. Ditto Bush the Younger against Al Gore and Ralph Nader. (The American Electoral College plays a role here which is different than the situation you describe.)


Leave a Reply

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

You are commenting using your 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

%d bloggers like this: