Serbs show French how premier-presidentialism works

Today voters in Serbia voted in a runoff election for the country’s presidency; legislative elections were held concurrent with the first round on 6 May. Meanwhile, France is in the interim period between presidential and legislative elections. What difference does this make?

France has long been seen as the model of semi-presidential government (notwithstanding that there actually are older examples). Specifically, it is of the premier-presidential subtype, which is to say that the president actually has very limited powers over government formation and policy-making, unless he leads a party or alliance of parties with a majority in the parliament. Under the premier-presidential subtype, the premier and cabinet are responsible to the parliamentary majority, but not to the president. Nonetheless, when the president is the acknowledged head of the legislative majority, he can be as unchecked in practice as any executive leader in any democracy.

The Serbian constitution, is unambiguously premier-presidential. Perhaps the presidency is very slightly less powerful, but the basic configuration of powers is similar to that of France.

So let’s compare the two countries, at this very moment, in terms of the process of government formation. In a premier-presidential system, “government formation” typically means the president initiates the appointment of a premier, but only upon taking account of the balance of forces in the parliament, which must approve his selection (and, solely, has the constitutional power to remove it subsequently).

In Serbia, the first round of the presidential election produced a close result, which was not decisive. Boris Tadic of the Democratic Party won the plurality, but only around a quarter of the valid votes. Close behind him was Tomislav Nikolic of the nationalist Serbian Progressive Party. In third place, but with only around 14%, was Ivica Dacic, of the Socialist Party of Serbia. (No other candidate had even 7.5%.)

This outcome made Dacic, the strongest of the candidates not qualifying for the runoff, potentially influential. To say “kingmaker” would be an overstatement, given that even if he could deliver his support as a bloc, neither candidate would reach 40%. Still, that did not stop some stories following the first round from suggesting Dacic would be the kingmaker.

Dacic tried, by announcing an alliance with Tadic, amid speculation that Dacic would become premier. Legislative elections were held at the same time as the first round, and they gave the alliance led by Tadic, called Choice for a Better Life, 67 seats. Nikolic’s alliance, Let’s Get Serbia Moving, won 73 seats. Dacic’s Socialists won 44 seats. With an assembly size of 250, a coalition led by Tadic and Dacic could combine for 111 seats–not enough for a majority, but with 44.4% of the seats, a strong base from which to build a government. Only one small detail: this coalition had to succeed in electing Tadic to the presidency first.

The voters did not cooperate, however, as Nikolic has won today’s runoff. Now Nikolic will need to begin negotiations to put together a cabinet that can command a majority in parliament.

This strikes me as more or less how premier-presidentialism is supposed to work. Parliamentary elections determine the parameters of coalition possibilities, given that–as in a parliamentary democracy–the cabinet must have the confidence of the parliamentary majority. Yet when there is no electorally based majority, it falls not to a third party in parliament, but to the voters, acting through their agent in the presidency, to serve as the real kingmaker.

Now contrast this process just sketched with that in France now. The presidential election is concluded, but parliamentary elections are looming in June. However, the newly inaugurated President, Francois Hollande, has already appointed his cabinet. Meanwhile, Hollande’s Socialists and the allies of the presidential candidate who finished fourth, Jean-Luc Melenchon of the Left Front, are divvying up the districts in which they will present joint candidacies, in order to maximize the seats of the broad left. In effect, Hollande (and Melenchon) are asking voters to ratify decisions they have taken since Hollande was voted into the presidency.

Events in France seem less in the spirit of premier-presidentialism, because they lend a far more “presidentialized” air to the whole process by permitting the appointment of the next government before the election of the parliament to which it is (formally) accountable.

The critical difference here is in the electoral cycle, with Serbia having its parliamentary elections concurrent with the first round of the presidential contest, whereas France, since 2002, has been employing a “honeymoon” cycle with parliamentary elections following close on the heels of the presidential runoff. When combined with the two-round majority-plurality system by which France elects its National Assembly, the honeymoon elections will tend to create a very large president-supporting majority, rather than a legislature that serves as a check on the president through coalition politics.

While both France and Serbia are clearly premier-presidential systems, the Serbian electoral cycle is much more in the spirit of the hybrid process of government formation that this subtype of constitutional form is supposed to generate.

17 thoughts on “Serbs show French how premier-presidentialism works

  1. The president-as-formateur model in Serbia sounds nice, but would it work as well with two-round legislative elections like France has?

    Serbia has (IIRC) quite a number of parties which are not necessarily aligned with others. A president’s party is likely to have at least some realistic prospects for a post-election coalition.

    On the other hand, the two-round system encourages parties to align into two large blocs before the election. This makes government formation much less comfortable when the president is from the parliamentary minority bloc. Either a party from the majority bloc has to betray its electoral partners from just a few weeks ago. Or, the president has to cooperate with the opposing bloc (“cohabitation”), despite the competition between blocs being zero-sum. It’s not surprising that under these conditions the French tried to get rid of divided government.

  2. I agree with your general argument, but two comments. First, the presidential inauguration in France was on 16 May, the second round of the legislatives will be on 17 June. So, there would have been slightly more than a month of cohabitation. Cohabitation may have benefits in some situations, but I am not sure that this would be one of them. A month is a long time in Euroland. Of course, the inauguration could be delayed or the elections held concurrently, but, as it stands, there is a certain logic to the French system. In Serbia, there is just two weeks between the presidential and legislative elections and no opportunity for cohabitation in the meantime. Secondly, it is at least possible that French presidents could lose the legislative election. Mitterrand returned a minority government in 1988. Currently, the UMP is polling higher than the PS, even if the PS has a lot more allies to call on in the 2nd round. So, ultimately, the legislative election will still shape government formation for the next five years. As you say, it seems to me that the big difference is the type of electoral system used for the legislative election in the two countries.

  3. I would regard the 1988 outcome as a success for Mitterrand, in that the election he called replaced a conservative majority. And Mitterrand called in the legisative campaign for an “opening to the center”. Maybe a minority government was not what he had in mind, but the outcome fits the pattern.

    I agree with both of you in the logic of the French pattern, particularly given the use of a non-PR legislative electoral system. But it is a far more presidentializing logic than what premier-presidentialism institutionally implies. (In that sense, it is, I admit, very de Gaullian, perhaps more than the old general could have imagined!). Whether it continues to be viable as a third force that doesn’t fit the 2-bloc pattern stabilizes despite the electoral system is an open question.

    On balance, I’d recommend premier-presidential systems go with concurrent or counter-honeymoon elections.

    Of course, if it’s true that the French pattern allows the president to ask the voters to ratify his government choices, then it is also true that the voters could refuse! But the electoral cycle makes such an outcome unlikely.

  4. I’ve seen the same exact point made a little early in the month on the internets, I think at the European Tribune site. Its a good point. Apparently when the French changed the presidential turn so the legislative and presidential elections would be in the same year, the original idea was for the legislative election to be held first. Voters would then have used the presidential election to either ratify the government selected in the legislative election, or to check it. Apparently a last minute deal between Jospin and Chirac resulted in the switch.

  5. One small question-the total of the two candidates in the second round don’t add up to 100%, is there a none of the above option in Serbia?

    The real unknowable in the French legislative election is the impact of the FN. Like 1997, the FN is expected to poll well (and may even win a seat or two this time round). However unlike 1997, the threshold for qualificiation for the second round has been raised from 12% of votes cast to 12% of voters registered.

    This translates typically into about 25% of votes cast, which may make it considerably harder for the FN to qualify for the second round, and thus make the contest between the centre-right and the united Left considerably narrower, as split votes on the Right would be less of a factor-the Presidential election confirming that a large majority of FN voters will vote for the mainstream Right in a straight contest between blocs.

  6. @Robert

    There is no period of cohabitation. Hollande already has his own prime minister and government in place.

    The tradition is that the prime minister resigns on the election of a new president. I am not at all sure of the logic behind that tradition. (We Westminster types have a strong tendency to say ‘convention’ instead of ‘tradition’!) Without that tradition the French presidency would be a much more standard premier-presidential office.

    MSS’ point is valid that France is an atypical premier-presidential system.

  7. I understood Robert’s point to be that if the French president waited until after the legislature was elected to choose a new cabinet, then there would be an interlude of cohabitation. And I do not think such an interim phase would be desirable.

    My point was that the ability of the president to change the government before the legislative elections makes the system especially “presidentialized”. But, as Samuels and I argue, we should expect premier-presidential systems generally to be fairly presidentialized. This point applies to Serbia as well: the president will now be in a position to lead the crafting of a coalition, and he is quite likely to succeed.

    In this point regarding presidentialization, France is more or less “typical” albeit more extreme since the change to five-year terms for both offices and honeymoon elections.

  8. I don’t think a newly-elected French president has ever lost their first (super-honeymoon?) election. Mitterand in 1988 had been president for 6 years.

    • There have been only five of these honeymoon elections in France, so a small ‘n’: 1981, 1988, 2002, 2007, 2012.

      And to emphasize a point I made above, even 1988 can’t be seen as an exception to the pattern. Mitterrand’s party went from 198 seats to 260, which was about 48%. Had he wanted to resume his post-1981 majority coalition with the Communists (24 seats), he could have. But he had called for an “opening to the center”, and the honeymoon election results made that possible, via the defeat of the incumbent right-wing coalition and the substantial boost for the reelected president’s party.

  9. Sorry if I wasn’t clear. MSS is right. I was presenting the hypothetical situation whereby Hollande was president but did not appoint a government.

    By the way, in Cape Verde there was an ‘interim’ cohabitation from Feb-March 2001 in the period between the presidential election and the inauguration.

    One last comment about Serbia. It’s worth remembering that Tadic resigned early as president. He could have gone on until February 2013. So, there were only concurrent elections because of his resignation. In that context, here’s another hypothetical. Imagine he had not resigned and the same parliamentary result had occurred. The outgoing coalition would have been returned and he would still be president.

  10. Alan,

    Mitterrand was newly reelected in 1988 when he dissolved parliament (elected in 1986).


    Threshold for participation in the 2nd round of legislative elections is 12.5% of registered voters since 1978.

  11. @Bancki, I had thought it was 12.5% of the votes cast, but perhaps I confused it with the cantonales, thanks

  12. In France, Hollande bolted totally the composition of the government. All the most important posts (economy and finance, education, social affairs, labour, home affairs, defense) are occupied by the main members of his campaign team. The only exception is the former PM, Fabius, as minister of foreign affairs.
    There is a coalition agreement between the Socialist party and EELV (the Greens). But there is no coalition agreement and no electoral agreement between the Socialist Party and the Left Front (Communist Party and Left Party of Melenchon) for the first round. The two parties did not succeed to agree about one candidate in some constituencies, where the Left has a risk of elimination after the first round, because of the threshold of 12,5 percent of registered voters. Moreover, the Socialist Party did not want to give “winning seats” to the Left Front. But there will be certainly the “republican discipline” for the second round.
    Pascal Perrineau, from Sciences Po Paris, gave a table about the recent honeymoon elections (Le Figaro, may 24, 2012)
    1981, F. Mitterrand (51,8 percent of voters) and Socialist Party (58 percent of seats, 285/491)
    1988, F. Mitterrand (54 percent of voters) and SP (47,7 percent of seats, 275/577)
    2002, J. Chirac (82,2 percent of voters) and UMP (63,2 percent of seats, 365/577)
    2007, N. Sarkozy (53,1 percent of voters) and UMP (55,5 percent of seats, 320/577).
    The rule of parliamentary confirmation of the presidential vote has an exception in 1988. There were minority governments (Rocard 1 and 2, Mrs Cresson, Beregovoy) during five years and the use of 49-3 article of the constitution

  13. The results of the first round of the legislativeelections in France seems to have confirmed the “honeymoon” effect.

    The French press are describing it as a good result but not a “pink wave”, since both the centre-right and the radical right have not collapsed completely. By contrast the radical left, while increasing its vote somewhat, may well be decimated in terms of seats, due to a better than expected scoialist performance in their traditional heartlands.

    It’s possible the PS will have a majority in its own right, but more probable that it will form a majority with its close allies, the Greens and the Left Radicals.

  14. Pulling this thread back to Serbia . . . the Serbian Deputy Prime Minister Kori Udovicki has announced a change of the electoral system, which appears to be a change to a mixed member system. She does not specify whether it will be compensatory or parallel. Does anyone know?

    • Considering that Serbia’s autonomous province of Voivodina uses parallel, I’m guessing that Serbia will have the same, though that is of course no guarantee.

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

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