France’s toxic combo of institutions finally bites

I still think Emmanuel Macron will win reelection, but it is going to be a closer fight than most prognosticators expected before this past Sunday’s first round. In the results of that vote, Macron has the expected plurality, and it was a few percentage points higher than he got in 2017 (27.8% vs. 24.0%). His runoff opponent in both 2017 and later this month, Marine Le Pen, also improved a bit over last time (23.3% vs. 21.3%). What is new–or really accelerating a trend that was already there–is the total collapse of older established parties. The Republican (mainstream right) got 20% in 2017 but only 4.8% this time, fifth place. The Socialists were already in dire shape in 2017 with 6.4%, but did even worse this time, 1.75%, despite (or because of?) running the mayor of Paris, a seemingly high-quality candidate. Jean-Luc Mélenchon of La France Insoumise, a far left group, made the race for a runoff slot pretty close this time, coming third with just under 22% (19.6% last time, fourth place). Given just over 7% for the far-far-right Eric Zenmour, one could say there was a majority for extremes of one sort or another.

While the Economist’s forecast model still has Macron’s win probability at around 80%, it was just short of 100% as recently as 21 March. An extreme right candidate actually has a roughly 20% chance of being the next president of France.

It is never a good thing for democracy when the fate of the republic hinges on one person. But it is hard to exaggerate how absolutely essential it is that Macron win. France has been running a decades-long experiment in whether a highly presidentialized system would eventually destroy the party system. The French party system held up pretty well, despite the adoption of a relatively strong presidency with the 1958 constitution and direct election to that office in 1965. The party system did indeed become presidentialized in ways that David Samuels and I document in our 2010 book, Presidents, Parties, and Prime Ministers. Parties reorganized themselves internally around the goal of advancing their presidential candidate, rather than emphasizing their parliamentary party organization. This presidentialization was only further enhanced by the decision in 2002 to make assembly elections follow immediately after presidential, with both elected for five-year terms. The party system’s left and right blocs, starting from the 1960s, came to be dominated by whichever party could present the successful presidential candidate–the identity of these parties changed over time on the right, but presidentialization allowed the Socialists to surpass the Communists on the left. However, with the demise of the old right and left, there is not much remaining to the party system other than presidential aspirations. Macron himself is the perfect demonstration of presidentialization–having no party at all till he was on the cusp of the presidency, and then creating one that swept into power on the heels of his own win.

The combination of direct election of a politically powerful presidency, honeymoon election of the assembly, and majoritarian electoral rules is toxic. It means that someone from outside the party system potentially can win the presidency and then, in short order, a majority in the assembly. If you get lucky with this combination, you get a Macron. If you get unlucky, you get a Le Pen (or potentially a Mélenchon).

Make no mistake. Honeymoon elections, with majoritarian rules, are the real deal. If Le Pen manages to win the runoff, there will be no “second chance” at which voters can check her with a majority opposed to her in a cohabitation via the assembly. Presidential and semi-presidential democracies just do not work that way. If she wins the runoff, we can expect her National Rally to win around 28% of the vote in the first round of the assembly (see the just-linked post or the one from 2017), and that to be a plurality. Could a broad alliance form to block her candidates, given the two-round majority-plurality system? Sure. Just don’t count on it. Do count on her getting support from various other anti-system forces and being in a much stronger position going into the second round of the assembly election than that 28% estimate implies.

Do I think this is the most likely outcome? No, I do not. I think Macron will win, and go on to win a large majority of the assembly. However, it is a bad situation for French democracy–and the world–to be dependent on this one man not slipping up in some way in the final days before the presidential runoff–especially with a major war going on in the extended neighborhood and related economic difficulties at home. France is in dangerous territory in these moments with its toxic institutional combo, and the overly high stakes that combo generates.

24 thoughts on “France’s toxic combo of institutions finally bites

  1. Macron proposed a “solution” to the Fifth Republic’s institutions problem in an interview today: Return the presidential term to seven years (from the current five), and take inspiration from the USA and hold midterms for the National Assembly.


      • Sure seems that way! That structure seems custom-built to lock-in cohabitation about half the time, and I don’t see how that’s better than the current set-up of cohabitation about never.

        On the other hand, maybe the mid-term structure might advantage the old government parties by making the midterm lower stakes? Like, maybe you have personality-driven parties/”movements” during the presidential election, but voters might behave differently when electing a mid-term government? Macron, Le Pen, and Mélenchon’s parties have done much worse in local elections in 2020 and 2021, which were still dominated by the centre-right and centre-right constellation of government parties.


      • Yeah the case for this model is that it creates quasi-parliamentary elections every so often which avoid the messiness that is two-round elections with single candidates (the effective SNTV race that is the contest for the places in the second round in particular).

        France is an interesting case for cohabitation largely because (as I understand it) the French Constitution vests most executive power in the Prime Minister, not the President, so cohabitation means that the Prime Minister is largely the one in control, so those midterm elections would be highly significant. It also means that the division of powers is seemingly relatively clean, without Sri Lanka or Ukraine-style constitutional crises (at least, past experience with cohabitation hasn’t led to these). Of course, the dissolution power means that a newly elected President dissatisfied with cohabitation can get a honeymoon election anyway, as Mitterrand did twice, and I don’t really understand the case for parliamentarism some of the time unless you regard the minimalism of the change as inherently a virtue.


    • Well, it would allow approximately 29% of the time to be under cohabitation, given the dynamics of late-term assembly elections. Maybe better than the status quo, actually, but not good.

      If there is only one reform they would consider and it is not PR, they should move assembly elections–both rounds–to be concurrent with the presidential election, or move them into the counter-honeymoon (i.e., a month or two before presidential rather than a month after).


      • Or perhaps they should adopt AV for presidential elections and allow the voters themselves to converge on two candidates. If AV had applied in the Chirac/Le Pen election, the final two would almost certainly have been Chirac and the Socialist candidate, Lionel Jospin. If no candidate reached the quota they could hold a runoff, but the runoff candidates would almost certainly be different from the current system.


  2. All French presidents like to have institutional reforms on their agenda. In 2018-19, Macron initiated his own constitutional reform package with a reduction of both houses of parliament in line with the cube root + a part of the Assemblée nationale elected by PR (discussed here earlier). He lost interest and did not finalise it when covid came along.


  3. Are you willing to dabble in some numbers?

    I think Macron’s chances of winning are something like 55-65%. No, one should not be reassured by 65%, even if it was the only number I gave – rolling a 1 or a 2 on a 6-sided die is a distinct possibility. The 80% proffered by the Economist is a bit silly (you’d think outlets like that would have learnt from 2016).

    However, I’m less certain than you that a Le Pen win would most likely mean a FN majority. I think it’s about 50-50, and I even think a non-FN majority slightly more likely than not. Let’s say 40-50% that the FN wins a majority. No, not reassuring, just not the most likely outcome. The dynamics of the second round for the assembly have always massively hobbled the FN, even in recent years.

    You said “Do count on her getting support from various other anti-system forces” Have other extreme candidates come out in favour of voting for her in the second round? Just curious.

    If you’re willing to offer your own numbers, I’d love to hear them!


    • I don’t have any basis for putting specific probabilities or ranges. What you propose, JD, is reasonable enough.

      As for behavior leading up to the second round of the assembly election if Le Pen has won, I do not think it matters as much what other party leaders are saying now, before the second round of the presidential election. What would matter most is what they would do after she is (hypothetically) elected.

      Please note also that I was careful not to predict a seat majority for Le Pen’s own party, only to suggest that a majority for the combined forces of those who voted for Macron in a losing presidential runoff would not be the most likely outcome.

      If she were smart, upon winning, she’d appoint someone not of her own party to be premier (I have no idea who might be available) and make various other moves to appear “reasonable.” Being president could immediately change the dynamic of second-round punishment that National Rally has faced before.

      If I did agent-based modeling, I’d give some post-MLP win scenarios a try! In the real world, I sincerely hope we do not get to find out.


      • Yeah I struggle to see a scenario in which FN doesn’t win a majority in the legislative elections. Just as was the case with Macron and LREM, there will be a not-insignificant number of LR politicians who decide not to stick with their obviously collapsing party and who would be very willing to be appointed to cabinet positions or nominated as FN/’Presidential Majority’ candidates. The extremist squeeze will probably be less significant this time if Melenchon’s party wins the lion’s share of the left-wing vote: if turnout is low, a bunch of second-round races will end up between FN and FI.


  4. “If she were smart, upon winning, she’d appoint someone not of her own party to be premier (I have no idea who might be available) and make various other moves to appear “reasonable.” Being president could immediately change the dynamic of second-round punishment that National Rally has faced before.”

    She certainly seems aware that this is a good idea: she promised to do this back in 2017. She said during that runoff campaign that she would appoint Nicolas Dupont-Aignan, a former UMP politician who now runs for President under his own banner every five years and broke the cordon sanitaire by endorsing her after the first round in 2017. She’s not making any similar commitments this time around (“we aren’t in the US, we don’t elect a ticket”).


  5. The combination of direct election of a politically powerful presidency, honeymoon election of the assembly, and majoritarian electoral rules is toxic. It means that someone from outside the party system potentially can win the presidency and then, in short order, a majority in the assembly. If you get lucky with this combination, you get a Macron. If you get unlucky, you get a Le Pen (or potentially a Mélenchon).

    Is that not roughly what has happened with the Trumpification of the Republican party? Except that Trump was able to capture the presidency on a minority vote?


    • Fortunately, Trump did not have a honeymoon election. I know the polls never showed much of a honeymoon effect, but he had a slight bounce, and it is hard to imagine he would not have benefitted had Congress been elected just a month or so later.

      So yes, it is quite similar. Except for the ways in which it is different. Comparative politics in a nutshell!


    • Matthew, you say with “the demise of the old right and left, there is not much remaining to the party system other than presidential aspirations.”

      And regional elections. In June 2021 the winning traditional parties in 2015 all retained control in 2021: “Republicans” (moderate conservatives) in seven regions, Socialists and their allies in five regions. Macron and his centrist allies were shut out of the 2021 results in three of the 12 regions, falling below the 10% threshold. In only one region, Le Pen’s stronghold of Provence-Alpes-Côte d’Azur, was Macron’s party part of a 6-party winning alliance of everyone but the Left against Le Pen. Example: In the conservatives’ largest region, Ile de France (Paris Region), the incumbent conservative alliance (headed by this year’s conservative candidate for President) got 36.19% in the first round. The left was split between three lists which got 34.36% in the first round, but merged for the second round and got 33.67%. The hard-right (Le Pen and allies) got 13.14% but this shrank to 10.79% in the second round to stop the left, just as the Macron centrists got 11.81% which shank to 9.62% as the conservatives jumped to 45.92% in the second round. With the 25% winner’s bonus, they got 59.8% of the seats.

      The regional elections had also left Mélenchon and his “France Unbowed” party with no momentum. In seven of the regions he had been shut out, including four of the five won by the left. In the five regions where France Unbowed got a few seats, he got slim pickings and no power whatever.

      In October 2021 Anne Hidalgo, mayor of Paris, won 72% of the votes in the Socialist Party primary. Some polls showed her tied with Mélenchon. On 4 December 2021, Valérie Pécresse, President of the Paris Region, won the Republican (conservative) nomination with 60.95%. Back before the polarization of “Le Pen vs. Macron,” the two women (Hidalgo and Pécresse) would have been the front-runners. Pécresse jumped into second place in the polls, ahead of Le Pen, not far behind Macron.

      What demolished the party system was the polls, the unofficial first round of the two-round system, really a three-round system. By March 12 Pécresse had dropped to third, tied with Mélenchon who then moved into third, rising as the Green and Socialist slipped. By March 31, Pécresse had dropped below 10% in the polls, and it was clearly a 3-man race.

      How the regional PR model hurt Mélenchon in 2021: his best region was Centre-Val de Loire, centred on Orléans, with 77 seats. Mélenchon shared 12 seats won by the joint list headed by the Greens. In the second-round merged list the socialists got 24 and the communists 6 in the governing caucus of 44. France Unbowed got no seat on the 13-member Executive, which is nine Socialists, three Greens, and one centre-left Left Radical, shutting out the far left.

      The winner-take-all tactics in the Presidential race over-rode the party system in 2017, but it bounced right back in 2021. I greatly doubt they have destroyed it this year either.


      • Fair points. The parties are still there. But that the system can be quite badly battered by what happens in the winner-take-all contest–in other words, presidentialization–is exactly the point. The older parties won’t fully recover in time for assembly elections, even if they bounce back for later regional elections.

        Just look at 2017 for example. After getting 6.4% in the first round of the presidential election, the PS got 7.4% in the first round for the assembly. So, yep, it went up. But that’s still pretty bad! (29% in 2012.) The Republican got 20% for president and the party got 15.8% for assembly. Quite a drop! (27% for UMP in 2012.)

        Parties whose candidates do badly in presidential elections do not typically make a comeback in honeymoon assembly elections. So I stand by what I said in the main post: Destroyed, but sometimes what is destroyed can be rebuilt–over time.


    • A “proportional” system that automatically gives 1/3 of the seats as a bonus to the largest party in terms of votes may actually be worse than the current system. The leading party needs just over a quarter of the vote to get a majority.


      • Indeed. Even the 25% bonus for regional elections is undemocratic. In only one of the 12 regions did it result in a minority government in 2021: Brittany, centred on Rennes.

        Brittany’s 3-party Socialist alliance got 20.95%. Merged in the second round with an anti-pesticide crusader, as well as picking up votes from the far-left France Unbowed’s 5.57%, they got 29.84%. This gave them the bonus, but still only 48.2% of seats.
        The Green+local autonomist list broke the 10% barrier this time, getting 14.84% in the first round and 20.22% in the second, and 14.5% of the seats. This allowed the socialist (moderate social democrat) Regional President to be re-elected, heading a minority government with 40 Councillors (28 Socialists, 4 Communists, 5 ecologists, and 3 regionalists) from 4 Departments (within each list, seats are allocated by department in proportion to the number of votes in each department, so they are more local than they look). They need support from either the 6 Greens or the 6 Breton Autonomists.
        To finish the story, a conservative list got 16.28% in the first round and 21.98% in the second (picking up some “stop-the left” switchers as Macron and centrist allies got 15.53% in the first round, but dropped to 14.75% in the second, no joy for Macron with 9 of the 83 seats). The conservative list got 16.9% of the seats. LePen and allies got 13.22%, and 9.6% of seats.


  6. Pingback: The French Election May Be Close – Variety Newz

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 )

Connecting to %s

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