How unusual is South Africa?

[Update: I added in a comment a list of cases with streaks of four or more elections with largest party having over half the vote.]

Given the news that a proposed merger of two opposition parties in South Africa was called off, I was wondering just how unusual is South Africa in having had a string of elections in which one party won a substantial majority of the vote.

Querying a dataset of over 1000 elections in countries that can be considered democracies (including many developing countries and young democratic regimes), the answer is: quite unusual.

The mean vote share of the largest party in these 1,148 elections is .42. In South Africa’s four democratic elections to date, the ANC has averaged .64. Such a high figure is just beyond the 95th percentile (.63) of the worldwide distribution.

Of course, the failure of opposition parties to offer a clear alternative does not help the cause of “normalizing” South Africa relative to other democracies. On the other hand, the extreme PR system also is an impediment. While normally, the more extreme the PR, the less likely is a party with over half the votes, if your starting point is a hegemonic party, then extreme PR presumably inhibits the development of an opposition that could challenge the leader. That is, numerous very small parties can all win their share without unifying, and the benefits from unifying are less clear for parties than they would be in a country with either a less proportional system or direct presidential elections (or both).

Speaking of direct presidential elections, the distribution of the worldwide data does not change much if we focus only on parliamentary systems, although there is small (and totally insignificant) tendency for bigger shares for the largest party under presidential systems (mean .43, although 95th percentile .61).

40 thoughts on “How unusual is South Africa?

  1. Can it be seriously contended that, if South Africa had retained the FPTP electoral system that prevailed under apartheid, the ANC would not hold just under 100% of the seats? The great problem in South Africa is not the electoral system, but that a large percentage of the African majority believe the DA would restore apartheid if they came to power. That is particularly ironic because the DA has its roots on the white liberal opposition to apartheid and figures like Helen Suzman who was, for many years, the sole opposition MP in the white parliament.

    The ANC hegemony is certainly reinforced by their ability to re-deploy people to the national and provincial assemblies at will, but not, I think by PR itself. The DA also has serious tactical weaknesses because its leader, despite impeccable antiapartheid credentials, is an Afrikaner the ANC can derisively dismiss as ‘Madam Baas’, the Afrikaans for Madam Boss and that leader is understandably more enthusiastic about retaining the premiership of the Western cape than taking a conventional role as opposition leader in the national assembly.

    • Of course, the ANC would have won a massive over-representation under FPTP. As I said, this is path dependent. Without a hegemonic party, and with a less proportional (not necessarily FPTP!) system, opposition parties would have more incentive to coordinate than they do under the current conditions. In that sense, PR and hegemony of ANC reinforce each other–which is rather contrary to the way we all normally view electoral-system incentives. It is an interesting case for these reasons (even before we get into specific features of the largest opposition party, or to the intra-party dimension).

    • And, unlike the ANC, the government in Singapore has been massively repressive in its treatment of political opponents. At one stage in the 90s all former opposition leaders were in jail or bankruptcy. The government makes a practice of having the courts declare political criticism defamatory and then recovering the ‘damages’ from the personal wealth of its opponents. They also use charges of criminals defamation to imprison and fine opponents.

      Really, Singapore belongs in the authoritarian column with Malaysia and the forms of state should regarded as window-dressing.

  2. The statistics I quoted yesterday do include some marginal democracies. However, if I repeat the procedure with only those elections under a regime with a Polity score of at least 5 (a commonly used cutoff), the distribution hardly changes. (There are just 139 out of 1148 in the larger dataset where Polity<5.)

    But for the time series question, one would want to consider only countries with sufficient level of democracy. If we use Polity=5 as "sufficient" then both Malaysia and Singapore drop out.

    Following is a list of streaks of four elections or more with largest share over .5 under conditions of Polity 5 or higher (includes some cases where Polity scores are not available, but case is surely democratic), showing years of streak and maximum largest vote share during streak, and some additional notes on some cases:

    Bahamas, 5, 1982-2002 (max .569)
    Barbados, 4, 1976-1991 (max .594)
    Belize, 5, 1989-2008 (max .597)
    Botswana, 8, 1974-2009 (max .766)
    –not over .55 since 1989 or over .67 since 1984)
    Gambia, 5, 1972-1992 (max .69)
    –ended with coup
    Guayana, 4, 1992-2006 (max .547)
    Honduras, 4, 1981-1993 (max .539)
    Jamaica, 11, 1959-2011 (max .601)
    –max actually .897, but that was a boycotted election (1983)
    –not over .557 since 1997, just .501 in 2007
    Malta, 7, 1976-2003 (max .518)
    St. Lucia, 13, 1961-2011 (max .619 in 1997)
    –was .51 in 2011; only twice over .6 (1961, 1997)
    St. Vincent & the Grenadines, 7, 1984-2010 (max .663)
    –not over .57 since 1994
    Trinidad and Tobago, 5, 1966-1986 (max .663 in last of streak)
    –was .841 in 1971, but that must have been boycotted
    USA, 4, 1984-1990 (max .541 in 1986)

    Many are in micro-states and few have more than a sporadic election above 60%; also I noted that in many there is a general (though not monotonic) decline over time. So, yes, South Africa is unusual. It is a big country that now has now four straight democratic elections with the largest party over .63, and that share actually grew over the next two elections (max .696 in 2004) before falling slightly (to .659 in 2009).

    Also note that some of those cases (e.g. Honduras, Jamaica, and Malta) have alternations during their streaks.

    (Countries enter this dataset around 1946 or when they become democratic, whichever is later.)

    • Namibia should be on that list as well, I believe. They’ve had a Polity score of 6 since independence, and SWAPO has won all 4 elections, with a minimum of .753 in 2009 and a maximum of .768 in 1999 (also .763 in 1994 and .764 in 2004).

      • Namibia is not in the dataset. I am not sure why (I did not construct the original on which it is based). And the latest elections included as 2012.

        Thanks for the additions/corrections.

        (The dataset starts off with Borman and Golder’s electoral-system database, and has been variously augmented by a student of mine. I suspect I will have more to say about our efforts in the near future.)

      • those are presidential scores. The legislative scores are scarcely different, however: .739 in 1994, .762 in 1999, .761 in 2004, and .743 in 2009. The streak would be 5 if one includes the 1989 Constituent Assembly elections, which SWAPO won with 57.3% of the vote

  3. The most recent electoral reform proposal, from the DA, would have made for 300 seats elected in 100 3-seat List-PR constituencies, with 100 national (presumably compensatory) list seats.

    • South Africa already has an established system of MMP for municipal elections. Why electoral reformers feel the need to reinvent the wheel (and often designate it ‘Spiffy Round Object Voting’) is beyond me.

      • Would MMP be such a great idea when almost all of the constituencies would be supersafe ANC-districts?

  4. The opposition has been able to win significant elections in SMDs, most notably in the Cape Town City Council. You’d have to be careful how you’d salt with overhang seats because the constitution itself mandates proportional representation. The supersafe ANC districts would be compensated by list MPs. obviously the overhang rules would be central to the fairness of the system.

    I suspect the ANC vote will go down significantly and there is a chance the DA will add the province of Gauteng to their existing hold on the Western Cape. However, the ANC is going to continue racking up landslides until their control of COSATU, the Congress of South African Trade Unions, breaks down or until a significant ethnic group breaks from their column.

    Formally the government is a tripartite alliance of the ANC, COSATU and the SACP, the Saith African Communist Party. At present an alliance of 9 ‘rebel’ unions, led by NUMSA (the metalworkers and the largest union) is demanding a special COSATU congress and the leadership has been stalling on theta demand for months. It is not unlikely that COSATU will formally split and we will see a labor-based party challenging the ANC from the left in 2019.

    • I recall a simulation that claimed that FPTP would not have a markedly greater ANC representation, contrary to what I acknowledged the other day, due to regional concentrations of opposition support. I don’t know if that is accurate, or still true now if it was valid in past elections.

  5. Incidentally does anyone know why municipal councils so large in the country? The Cape Town council has 121 district councillors and 120 list councillors.

  6. What I’m really concerned about is not overrepresentation of the ANC, but rather district seats that would be so safe they wouldn’t be too different from their current safe list positions, offering few incentives for better representation and very little effective choice in the election. But perhaps I’m exaggerating the problem.

    • I do not know if it’s within the DA proposal but there has been a lot of discussion about MMP and making district MPs independent of their parties, i.e. not subject to redeployment.

      • Sounds interesting, but I’m not sure I understand. What exactly would that entail? Removing party labels from ballots?

  7. The South African constitution provides that you lose your seat if you cease to be a member of the party which nominated you. The ANC has developed this principle into ‘redeploying’ its members from, say, the national assembly to a local council. It is not a system that encourages a high degree of independent action among MPs or municipal councillors.

    The ANC was somewhat terrified by the vents in a small municipality in Northwest province named Tlokwe.

    Dissident ANC councillors voted in favour of a DA motion of no confidence in the mayor and then helped elect a DA mayor. The ANC immediately redeployed its list councillors out office but had to win by-elections for the district councillors to regain its majority. Sadly for the dissident councillors the by-elections were held the week after the death of Nelson Mandela and 5 of the 6 dissidents were defeated and the ANC regained control of the council.

  8. What about an open party list system in South Africa? It seems to me that would be better than MMP system since there would be no need for overhang seats.

    It seems likely that the ANC will rule South Africa for a very long time, so it is a democratic dominant party system similiar to India’s Congress Party (FPTP), and Japan’s LDP (SNTV, since 1993 MMM), but has any party ever win majorities that most parties dream of, all with the most proportional system of proportional representation. This is very unusual. Most South Africans like actually love the ANC, so they vote for it. I think for most democracy advocates, a dominant party system is not what they think of democracy as they want alternation in power between parties. Fortunately South Africa is stable, and not like Bulgaria (no government has ever been reelected), and Bangladesh where there is alternation between parties, but can one really say that these are good democracies?

    It seems very surprising considering that the ANC is a Social Democratic party that allies with groups that are to the left of it. It seems to me that South Africa is lacking a Center-Right Conservative Party for the Black Majority, would the Inkatha Freedom Party ever someday fill this gap.

  9. Rob, in one critical respect, South Africa is like neither India nor Japan in their dominant-party eras: the ANC is far more dominant. The Indian Congress never won close to half the votes, and the Japanese Liberal Democrats only occasionally did, and never close to 60%.

    • One of the nastier facts about SA politics is that a majority continue to believe that apartheid would be restored if the opposition took power. The ‘born frees’, people born after 1994 have significantly different opinions but it will take time for them to become a dominate force in the electorate.

      • Last time, some people who can’t remember apartheid could vote for the first time; in the upcoming election, some people who were born after apartheid can vote for the first time.

    • How about Malaysia, but then Malaysia has an authoritarian dominant party system? Is there any other examples of countries like South Africa where a party is as dominant as the ANC? I can think of most of the rest of the Southern African countries; Namibia, Botswana, and Mozambique. What is the cause of this development?

      South Africa’s opposition doesn’t have the excuse of blaming the electoral system for their incompetency like the Indian, Japanese, and Malaysian oppositions because of PR.

  10. Subnationally you have Alberta and Bavaria, but even then often under 50% of the vote, and only rarely over 60%.

  11. Malaysia is not a democracy, according to Polity or Freedom House. Mozambique is, but for whatever reason is not in the Bormann and Golder dataset, which was our starting point for this analysis.

    I suppose the dominant parties in various southern African cases is simply a product of the “national liberation movement” phenomenon. Mozambique, Namibia, and South Africa are all PR systems, but all have a party that emerged dominant in the struggle against colonial or apartheid rule. I suspect either single-party rule or democracy will have to give way at some point, but who can say when, or even if?

    • Is Malaysia and Singapore the borderline case of what is not a democracy? In both of these countries, the opposition could theoretically win an election, then they would be consider a democracy after that election, but not before?

      • Rob, that’s a good question that may not have a clear answer. Can you (or anyone) think of a good example of a case generally classified as not democratic, but with elections, where the opposition won and took power without an irregular process (e.g. the military rebellion that enforced the rightful winner in the Philippines, 1986)? I have been thinking of this off an on today, and have not come up with an example.

        I guess–and this is all it is–that if such a thing happened, Polity would classify the period after the election as “in transition” while waiting for the dust to settle and to make sure the newly elected government repealed anti-democratic legislation and did not simply reinforce the old limits on democracy. But I just can’t think of a good example of such an actual case.

        And, in case anyone is thinking of Mexico, I believe most sources count it as democratic already by 1997 (some would say earlier), not only when the executive changed with the 2000 election.

  12. Puerto Rico’s Popular Democratic Party (PPD) won landslide victories in every general election held in the Island between 1944 and 1964; the party, which came to power after a closely fought race in 1940, went on to win 64.7% in 1944, and in the next five general elections its share of the vote fluctuated between 58.2% and 64.9%.

    By the mid-1960s some political scientists were actually writing about the PPD becoming “institutionalized” (much like Mexico’s PRI), but a damaging split in 1968 led to the party’s defeat after twenty-eight years in power, and the Popular Democrats, while alternating in office since then with the pro-statehood New Progressive Party (PNP), have never been able to score another landslide victory of the same magnitude.

    • Manuel, thank you for that example. Puerto Rico is too easily forgotten by us comparative-politics scholars. It falls into that grey area between sovereign state and “others that we usually don’t include” (for lack of a better category name). It is in some comparative databases (like the Constituency Level Electoral Archives), but not in most.

    • That is definitely a democratic dominant party system. PPD ran Puerto Rico for over 25 years, and that is a dominant party system era. It ended. Puerto Rico has alternation in power between the two parties. I wonder if one can say the same about another country with a dominant party system Sweden with the Social Democrats. Sweden like any Scandinavian country are known as being the pinnacle and the paragon of democratic virtue. It has been out of power for now 8 years. The longest period that it has not ran Sweden.

      I know I am being sarcastic tongue in cheek. Can any country be lauded as being the pinnacle and paragon of democratic virtue? Or can no country be a perfect democracy? Is democracy like perfection, it goes all the way to infinity that never ends?

  13. Let us bear in mind that the national assembly and all 9 provincial legislatures face a general election this week.

    • It occurs to me that this is an interesting test to the question of whether concurrent elections can lead to more split votes and therefore opposing forces in the provinces (and in the NCOP).

      • Are concurrent national and provincial elections unusual in South Africa? Wouldn’t we generally expect concurrence to dampen split-voting?

    • No, this is of course nothing new. And indeed concurrent elections usually have that effect, but perhaps not in this kind of context; as the ANC slowly weakens, I think there might increasingly be a chance of vote splitting of the kind that happens in Australia to some extent, as the provincial assemblies elect the upper house. But this is nothing more than speculation.

      • The constitution binds the provincial legislatures to elect NCOP delegates by proportional representation.

        I’d be surprised if the DA does not retain control in the Western Cape. I suspect the opposition has a chance in Gauteng and Northern Cape.

        They may have a remote, remote chance in Limpopo where the ANC performance has been particularly dreadful and two opposition party leaders, Julius Malema and Mamphele come from. Mamphele was briefly the DA presidential candidate although she leads a different party. Malema was sacked by the ANC as head of their Youth League some years ago.

        South Africa strikes me as a very good argument for federalism, even though the provinces are significantly weaker than in most federations. The opposition would be profoundly weaker were it not for their ability to chip off provinces one by one and to point to their fairly spectacular record in the Western Cape.

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s