Egyptian coup threat–and a proposal for a presidential-election method

Following a weekend of remarkable protests against President Mohammad Morsi, Egypt’s armed forces have made an ominous declaration:

Wasting more time will mean more division and conflict, which is what the armed forces warned of and of which it continues to warn.

In the statement, the army gives the civilian leadership 48 hours to “fulfill the people’s demands” or else it will offer a political “roadmap”. Despite stating that the military “will not become involved in politics or administration”, the statement sure seems like a coup threat.

The failures of Egypt’s transition from the Mubarak era to something different, whether democracy or whatever, have roots in areas I am not intellectually competent to address. It is certainly a deeper problem than bad institutional design and sequencing. However, Egypt is quite the textbook example of bad institutional design and sequencing. First a few reminders on the sequencing, then some thoughts on an institutional design that might be too late for Egypt, but potentially helpful in some future case of transition from dictatorship.

Egypt has an elected presidency but a non-functioning legislature. The specific occasion for the current protests is the first anniversary of Morsi’s assuming executive power. The legislature’s first chamber was dissolved following a judicial decree that the electoral system was unconstitutional; the electoral system itself was indeed poorly designed. Let’s just say that this crisis of civilian authority is not exactly a surprise, given this track record. ((And there were warning signs even earlier, given the unusually low turnout in a transitional referendum.)) I am not aware of any successful transition to democracy in which there was an elected presidency in the absence of an elected legislature–the advice of at least one US political scientist that Egypt should set exactly such a precedent notwithstanding. ((I am not sure of the current status of the second chamber of the national legislature. Even if it functions, the absence of a first chamber is itself symptomatic of a dearth of checks on the presidency.))

But the president has a popular mandate, so what’s wrong? Well, not much of a mandate. Egypt, like many presidential (and semi-presidential) countries, elects its presidency by two-round majority. In more institutionalized settings, this might be a perfectly fine rule. But in transitional settings, it often may not be. The top-two rule, by requiring a second round if there is non consensus in the initial vote, ensures that the winner will have a majority of votes cast, and thus in principle prevents an “extremist” from winning. However, the achilles heel of this election method is the first round. What if the top two finishers in a crowded field are both “extreme” in some sense? For instance, one heads an ideological movement that was not the prime mover of the anti-dictatorial protest, while the other was a holdover from the dictatorship itself? What if these top two did not even represent half the votes cast in the first round? As I said at the time, the Egyptian first-round result was “Not promising.

In fact, a more moderate/liberal candidate finished a close third, and was thus eliminated. He might have been a Condorcet winner, meaning the candidate who would beat any other in one-on-one competition, but of course, he never got the opportunity to show if he was. There were other relatively liberal candidates in the race as well, which of course is the problem: they split the vote, whereas the Islamists (Morsi) and old-regime forces (runner up Ahmed Shafiq) were more unified.

This leads me to my modest proposal. To guard against divisions in an unsettled political field during a transitional process, why not take a page from the old Uruguayan system of presidential election? In Uruguay, it used to be the case that the presidency was chosen by essentially an open-list PR system. A party or alliance could have several candidates, and the winner was the candidate with the most individual votes within the party/alliance with the most collective votes. I am not suggesting this system exactly: I think either a majority-opposed president or a candidate with the most personal votes who nonetheless loses (because he was not in the biggest party/alliance) is a potentially dangerous idea in cases of poorly defined political lines–such as Egypt currently. Rather, I am proposing such a system for the first round only. The eventual winner would still require a majority.

Here is how it would work: candidates would have the option of declaring alliances with other candidates, once registration of contenders was complete. If one of the candidates won an absolute majority, the race would be over, and the alliance declarations would not come into play. However, in the event no candidate won a majority of votes, the runoff pairings would not necessarily be the top two candidates. Instead, they would be the leading candidates from each of the two largest alliances.

Of course, there is no guarantee that candidates representing a potential bloc of moderate voters, such as Egypt’s liberals, would agree to set aside their egos or policy disagreements and declare an alliance. But such a rule would give them the option, and at least provide the more moderate forces a chance of avoiding a catastrophic coordination failure. Such failure is inherent in a first round that is more SNTV (multiple candidates from a bloc or proto-bloc, but only candidate votes matter in setting the top two) to one that is more open-list (pooling of allied candidates’ votes being the first criterion in setting the runoff pairing).

One might argue that it would be better to use ranked-choice ballots, such as the alternative vote (a form of “instant” runoff). Or not to have a presidency at all. These might be good arguments. But actual transitional cases often have a presidency, and in practice never have ranked-choice ballots. I offer this modest proposal as a more palatable one–or so I think it could be–that addresses the basic problem.

It is too late to save Egypt from a major political crisis, and it is in any case a lot to ask of a presidential-election method to save a country from crisis. But at least the rules chosen should stay out of the way, and the top-two majority-runoff rule did at least fail to do no harm.

21 thoughts on “Egyptian coup threat–and a proposal for a presidential-election method

  1. “Former deputy head of Egypt’s General Intelligence apparatus Hossam Khairallah … said the legitimacy of Morsi as president had expired, and that the 48-hour deadline set by the armed forces was intended for him to prepare his resignation speech.” (Al Ahram)

    OK, so we are passed coup threat, and on to actual coup. Or at least coup attempt.

  2. I noted that Egypt was not the only case my modest proposal might be useful for. Ivory Coast comes to mind immediately as another one with a highly problematic first round. However, there, unlike Egypt, there were three candidates well dominant over the rest of the field.

  3. Outside the transferosphere (admittedly there seems little prospect of any need for transitional governance in Australia, Ireland or Malta) this would be a good rule. I am not sure about longterm and it would be good to avoid South Africa’s unhappy record of keeping a good transitional voting system long past its use by date.

    After the Egyptian case, we can hope no-one will ever again argue for an elected president before an elected legislature.

  4. I am not sure whether the rule (which remains nameless at this point) would be a hindrance in the longer run. If parties decided they did not want to join forces, they wouldn’t have to. However, I was thinking of it as a transitory provision.

    Of course, the only way to make it transitory is to have a sunset clause: a new rule–or an extension of the transitory one–required after a certain number of elections. I actually think sunset clauses on transitional democratic institutions are a good idea in general. Not only South Africa is an example of keeping an electoral system past its use-by date, but Israel!

    I, too, assume–and hope!–that advocates of elected presidencies in the absence of elected legislatures have reconsidered by now.

  5. The main advantage over IRV/alternative vote appears to be that it doesn’t require something people aren’t used to, in this case ranking candidates. Instead, this unnamed proposal supposes that candidates will know (or learn very quickly) how to form coalitions. During the transition from autocratic rule to something more democratic, are fledgling politicians any more used to that than voters are to ranked ballots?

    Assuming the answer to my first question is “yes”, parties and candidates have to think about whether coalitions are always in their strategic interest. My intuition is that the strategic considerations will not be as straightforward as just preventing the bad results of vote splitting. I’m sure it has happened that candidates perceive the blocking of someone who the voters think is close to them as more important than blocking the election of someone the voters think of as at the opposite end of the spectrum.

    I think this point would need careful thought (and maybe some analysis by the social choice theory/game theory folks) before the proposal should be implemented. Especially implemented in inherently unstable situations.

  6. Other idea could be, instead of only the two candidates with most votes passing to the second turn, the secund turn could be with the most voted candidates having 50%+ between them.

  7. The problem in Egypt is the collapse of authority. This happens more often than one might imagine and it seems to me there is a case for a heavily-safeguarded UN agency to handle these things.

    East Timor went through periods without government twice last century.Once in 1975, when the Portuguese simply lowered their flag and sailed away without designating anyone to assume control. The second occasion was when Indonesia in its turn abandoned the territory although its armed forces executed a genocidal massacre before leaving,

    The territory was governed by the UN Transitional Administration for East Timor from 25 October 1999 to 20 May 2002. Initially the UN high representative held all legislative and executive power while the courts were staffed by overseas and local judges appointed by the high representative. In due course power was devolved to a constituent assembly and a government was formed under UN supervision before the final grant of independence.

    Cambodia also went through a period of UN administration after the civil war.

    A similar,although more complex arrangement prevails in Bosnia-Herzegovina and in Kosovo, where the UN high representative retains a veto over legislation and the power to dismiss elected and unelected officials.

    There is probably now a majority of Egyptians who want the Mursi presidency dissolved. It is a pity their only option is the army rather than petitioning for a UN mission.

  8. Well, if there is such a majority, it apparently got its wishes fulfilled today.

    Miguel, I have thought of proposals like yours, as well. Of course, it would mean either a third round or accepting sub-majority winners in many cases.

  9. Obviously what has happened is a coup although there is some evidence that it has popular support. It still remains a coup even where the electoral system was not particularly effective and the president seemed quite comfortable governing without a parliament.

    Frankly, there is quite a lot of available expertise on transitions and it’s pretty amazing that the Egyptian authorities seem to have managed to choose the worst option every time they had to make a choice.

  10. A ranked-choice ballot would be a good proposal for a democratic government. Something as simple as a truncated rank 3 only idea would work if it has a twist saying that voters may give an equal preference to more than 1 candidate but less than say half of the total number of candidates.

    Example given:

    2012 U.S. Presidential Election

    Obama (3)
    Romney (3)
    Johnson (1)
    Stein (2)
    Goode (2)
    Anderson (2)

    An equal preference would give the candidates ranked a “shared vote”, equal to 1/n (where n is the number of candidates ranked here).

  11. Perhaps an indirectly elected President would be better in these circumstances-like the old Bolivian model, where the top three candidates in votes would have a run-off in the legislature-the legislators themselves elected in closed lists under the name of the candidates. (A candidate that won over 50% of the votes would win outright). That way a politically scattered bloc could reunite in the legislature and at the very least act to block a completely unacceptable candidate?

    • DC and Tom, fair points. Of course, for that, one needs an actual sitting legislature!

      In the Chilean example, I have pointed out before that we can’t be sure if it was a norm that led the legislature, in practice, always to “ratify” whoever won the most votes when that was below 50%, or if there was never a situation, prior to 1970, in which a majority of the legislature actually preferred the second candidate. In fact, 1958 was the only other time when the plurality winner was far short of 50%, and that year the number two was… Allende.

      As for the 1970 election, it was held at a time when the old right and the Christian Democrats (PDC) were vying for the rural vote, and the previous (Frei, PDC) administration had been very exclusionary in its dealings with the old right. The PDC may actually have preferred Allende over granting the presidency to their old nemesis on the right. Congress required Allende to sign a (rather humiliating) statute of guarantees of constitutional procedure as a condition for getting their presidential votes. That very act suggest it was bargained, and not simply accepted as a norm that Allende would get the congress’s votes.

      Anyway, back to Egypt…

  12. One idea worth considering, if runoffs are used, is to set the threshold for outright election as a percentage of the total eligible voters (rather than of votes cast). For example, if none gets over (say) 40% of the total electorate (whether because of a split vote, low turnout, or a combination of both), then the legislature (ideally elected by PR) chooses between the top two, and any other candidate who was supported by at least (say) 10% of the total electorate (to avoid a France 2002 situation). Multiple ballots if necessary but use the “lowest one out” rule if no candidate withdraws, you don’t have protracted deadlock.

    Chile, of course, formerly used legislative runoffs but they developed over time into de facto first past the post, with it being perceived as illegitimate for the Congress to reject the “winning” (sic) candidate – even in the 1970 Allende situation, where following the convention required right-wing Deputies and Senators to (in their view) hold their noses hard and ratify a presidential candidate they strongly opposed, elected on a comparatively low percentage of the votes.

    I wonder if the Shugart/ Samuels research on separation of purpose in presidential systems would lead to the conclusion that 50% in the legislature is a poor proxy for 50% in the populace when deciding whom The People want as their president?

  13. Particularly in transitional situations, you want rules to be simple. When, against expectations, the Indonesian electoral college chose Abdurrahman Wahid over Megawati Sukarnoputri, the country was very lucky that Megawati was so hopeless at running a We Was Robbed campaign.

    Trust is not very high in these situations and a transitional legislature electing a 10% candidate over the top two would just tend to confirm popular expectations of corrupt electoral processes. Allowing the parties to form coalitions, as MSS suggests, is a much more transparent approach if there is to be an elected transitional president.

    It may be that the best approach is for the transitional president to be appointed by a PR legislature on the Indonesian/South African model.

  14. Noted, MSS. Thomas Hauser’s Missing (1982) claimed very insistently that before 1973 almost all Chileans, apart from a few far-Right holdouts, would have considered it blatantly improper for the Congress to elect the second-place candidate over the plurality winner. Hauser even drew an analogy here with the US Electoral College — this was, of course, 18 years before-Bush v Gore … Maybe Hauser was exaggerating here; unnecessarily, if so, since the disappearance of US journalist Charles Horman (and of thousands of Chileans) is outrageous enough without distorting the constitutional history of Chile to add one more charge to the indictment against Pinochet and Kissinger (just as Watergate is bad enough even if one concedes that Nixon polled a large majority in 1972).

  15. The problem with these unwritten chunks of constitution is precisely that they are arguable.

    It’s more than a generation after the Whitlam dismissal and you still find heavily argued papers that the actions of the governor-general were or were not constitutional. Even the Gillard/Rudd transition has slight controversy because Gillard advised the governor-general to send for Rudd and most scholars seem to agree that an outgoing prime minister, especially one without the confidence of the parliament, cannot tender advice to the governor-general except for a dissolution.

  16. This may be a dumb question, but if an outgoing Prime Minister without confidence cannot advise the Governor General, how can the Governor General send for the leader of the other party after an election? Somehow I doubt that the GG is able and expected to pick the Prime Minister on his own. Surely even a defeated Prime Minister has the power to advise calling on his obvious successor.

  17. The governor-general’s most important job is selecting a prime minister who enjoys the confidence of the house of representatives.

    The prime minister has to be responsible to the parliament for every act of the governor-general. An outgoing prime minister cannot tender advice because there is no mechanism by which they can answer to the parliament for that advice. An incoming prime minister assumes responsibility for their own appointment and answers to the house for the appointment.

    The opposition had their chance to test Rudd’s government by moving a vote of no confidence. They failed to do so in the face of public statements from Katter and Wilkie that they would extend confidence to the new prime minister.

    The governor-general has published the Rudd/Gillard correspondence. You’ll note that Gillard not only gives advice she is not entitled to give but actually tries, equally improperly, to make her resignation conditional on the Rudd appointment. Rudd does better by offering to meet the house as soon as possible to give the opposition a chance to test confidence.

    It’s unfortunate the governor-general has not posted her replies to Gillard and Rudd.

  18. My bad. I was under the impression that the GG was only empowered to act independently in extraordinary circumstances–decidiing whether to go to the polls or summon the opposition after a no confidence vote, the Dismissal in ’75, the Tasmanian situation where a Governor was advised to call a person he believed clearly did not enjoy the confidence of parliament, etc.

  19. It’s a bit weird but appointing a prime minister is not a case of independent action because the newly-commissioned prime minister has to take responsibility for the appointment.

    Reserve powers only come into play when the governor-general acts without advice and without a first minister to take responsibility for the decision. So 1975 yes, refusing a dissolution yes, but the Tasmanian governor’s decision to commission the Labor leader no.

    We can assume the governor-general required Rudd to meet the parliament at the earliest opportunity but we won’t know until all the correspondence is released. The legal advice she sought from the solicitor-general is on the public record. Notice that she did seek her own legal advice and clearly believed she was not bound by the outgoing prime minister’s advice.

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 )

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