One last Senate election article – I promise.
The last time Series 1 of the French Senate was up for election, on 25 September 2011, I wrote an article explaining that the renouvellement represented the first act of a long electoral process that would culminate in the 2012 presidential and general elections.(1) With a pleasing symmetry, the renouvellement of 24 September 2017 closes another long electoral season. French voters will not go to the polls again until 2019 for European elections and the spring of 2020 for les municipales.
The contrast between what happened last Sunday, however, and that other September Sunday six years ago could not be more striking. Then, the broad left (Socialists, Greens, Left Radicals and their allies) went into the renouvellement as a more-or-less organised bloc. The municipal elections of March 2008 had acted as a vote sanction against the Sarkozy-Fillon administration and already in that year’s partial renewal, the left had made significant strides towards disturbing the right and centre-right majority in the upper house. Senate-watchers, myself included, anticipated that the upper house would fall to the left in 2011. How smugly right we were.
In 2011, the left won pretty much all the seats it had hoped for, took the majority and elected as its speaker the leader of the Socialist group in the Senate, Jean-Pierre Bel from the Ariège, a department down in the Pyrenees with a long left-wing tradition. Over the next two Sundays ‘le peuple de gauche’ chose François Hollande as its candidate and the rest is history. Hollande beat Sarkozy to the Elysée and the PS and its allies took a majority in the National Assembly.
For the first time in the history of the French Republic, the Socialists controlled the presidency, both houses of parliament, almost all the regional assemblies, a good chunk of the departments and most of France’s larger cities… How they contrived to throw it all away is a story for another day. In this piece I’d like look closely at the results from 24 September by means of some case studies.
The link to the results of the elections is here. As I’ve already mentioned in my previous two posts, the election involved 170 seats in the series, plus one by-election in Savoie. In electoral terms, the Senate is a mixed assembly. For departments with three seats or more, a system of highest-remaining average PR is used, in those with one or two, a straightforward majority system, with a run-off is used. On the map, the lighter departments in mainland France and in the overseas territories are the ones using the majority system.
Here’s how it works. Candidates in two-seat departments can make common cause, if they wish, and stand on a joint ticket, but there is no compulsion on the part of electors to choose them both. In France, electors place a slip with a candidate’s name printed on it into an envelope and then place that in the ballot box. So it is perfectly possible for any given elector to vote for, say, a Socialist and a right-wing Republican should they so wish. If any candidate gains an absolute majority of votes cast, he or she is elected at the first round. If one or two seats remain undecided, then the second round takes place after lunch, which can very often become a moment of intense bargaining between candidates and electors, especially where there are groups of delegates from the larger communes in the department who might want to leverage some sort of deal...
Because of the way the system works, there is no attempt to enforce parity legislation here, unlike the departments electing three senators, where a list system of PR with no transferable votes is used. In these departments, the law requires the candidates to be listed alternately by sex, though oddly enough, most têtes de liste are men…
Now, one might think that PR is a more ‘democratic’ system, less open to abuse (whatever that is) and that where a party or parties put up one unified list, they will do better. None of these things is necessarily true. [On a point of information, in 2011 PR applied to departments of 4 or more senators. The Socialists reduced it to three or more ahead of the 2014 renouvellement. This threshold has been altered at least twice since 2000, depending on which side is in power: lowered by the left, raised by the right.]
What always needs to be borne in mind when thinking about Senate elections is the matter of ‘district magnitude’ that is to say the number of electors, which is in fact very small. In series 1, the largest departmental college is the Nord (59), with 5856 electors choosing 11 senators. The next largest is neighbouring Pas-de-Calais (62), with a shade over 4000 electors for 7 seats, Seine-et-Marne (77) with 3216 for 6, and then Moselle (57), with 2926 for 5. Compare that to Paris (75), which has 12 seats but fewer than 2969 electors. The Yvelines (78) has roughly the same number of voters choosing just 6 senators. And so it goes.
Paris is exceptional because of the way its local government is divided up. The size of the colleges is not driven by the population of the department, but by the sliding scale of municipal delegates that I described in the first of this three-part series of entries.
There are also other, much older factors at work. When we look at the number of communes in a given department, it reveals ancient patterns of settlement. For example, the Nord has some 650 communes, the Pas-de-Calais more than 800. Other departments with similar populations have far fewer. There are only 119 communes in the Bouches-du-Rhône (Marseille) and fewer than 300 in the Rhône (Lyon). For a more extensive discussion of these phenomena, I refer the reader to my 2009 book on the Senate.(2)
Okay. Points taken, but how does the election work?
Well, in the first instance, the district magnitude, or rather the number of valid votes cast is divided by the number of seats available – a quotient – so that any list that has that number of votes or multiples thereof gets that number of seats. For example, in the Yvelines, Gérard Larcher’s department, his list ‘Une équipe pour toutes les Yvelines’ (LLR for Liste Les Républicains in the table below), won 1588 votes, a shade under 58% of the 2741 valid votes cast. (Senate electors can spoil their ballots. Also, electors must attend the vote, on pain of a fine if they do not have a bona fide note from their mum). The quotient, then, is 2741/7, which is one seat per 457 votes. Larcher’s list took 3 on that basis. None of the other lists would take any in the first instance. Then the highest remaining average system kicks in – the famous d’Hondt formula V/(s+1), where V is the number of votes and s the number of seats already attributed. Using this system the vote in the Yvelines turns out this way.
In the table I have only included the results of the four best-placed lists. After the attribution of three seats using the quotient, the others are handed out using the formula. Thus, LLR took the first four seats, LRM the fifth and LLR the sixth, comfortably ahead of the Socialist list, with just 245 votes.
The success of Larcher’s list and of many others was largely due to running single unified lists. But it isn’t always the way in Senate elections, where the district magnitude is really quite small and securing a seat might mean that a list need only secure three or four hundred votes - or in the case of Paris, 180. Sometimes a party can do very well by putting up several different lists.
This is particularly true in departments where there a significant number of supplementary delegates, my ‘mayfly’ electors, who are, to all intents and purposes, co-opted for the election by the parties within a municipal council and in proportion to each group's size. An influential figure within a city, town, or district of a department can be very effective at delivering voters to a candidate - or may even be the candidate.
Let’s look at some cases where division did little harm or in fact boosted performance..
Moselle (57) is a department in northern Lorraine, carved out of the German victory in the Franco-Prussian war of 1870-71 and the Allied one of 1918. It’s quite a populous department with five seats. Generally, it has been to the right, in Senate terms, although in 2011, the right took three, the left two. Among the outgoing senators, one of these left-wing senators had defected to LRM. The 2014 municipal elections went well for the right and one might have expected them to make common cause to maximise their chances. Not a bit of it. Instead, the right ran three different lists and took four of the seats, with the PS holding on to one and LRM losing theirs.
In the table above, MA stands for Moselle Avenir, a non-aligned right-wing list, LLR the official Republican list, LUG the Left Union, and NDM Une Nouvelle Dynamique pour la Moselle, another ‘independent’ right-wing list. And, as the table shows, their divisions did no harm to their cause. And the presidential candidates on the La Moselle en marche list managed only 242 votes. And it’s worth pointing out straightaway that a unified list would have made no difference. Even if we added three right-wing lists together, the left would still have taken the ‘third’ seat, thus:
The Manche (50) is the department that comprises the Cotentin peninsula, the bit of Normandy that sticks up into the English Channel and has Cherbourg at the top. It’s a largely rural department and has three seats. Here, the Republicans did not even try to unify their official lists and let two run. And won seats with both ahead of the left-wing list, which in turn pushed LRM out of the seats.
Here, however, assuming all the voters would have backed both lists (which might not always be the case), the LLR would have taken all three seats if they had stood together. What the figures in this instance do not tell us, however, is whether the head of the second Republican list, the outgoing senator Jean Bizet, refused to accept third place on a joint list (parité oblige) and suspected there might be a risk to being the third candidate. Since parity legislation was introduced for the 2001 Senate elections, it has been noticeable that some (many) male candidates, forced down the list as a consequence, prefer to run their own. Odd, that.
There are breakdowns of each election in each department available on the Le Monde website, or in the paper version for 26 September, but no account of Series 1 is ever complete without the extraordinary spectacle that Paris provides every time its senators are up for renewal.
Macron and his marcheurs did particularly well in the capital and, with so many supplementary delegates, the election is always a little unpredictable. Not only were there a record number of lists (13) chasing the 12 seats, but eight of them got at least one. As Le Monde put it in its election supplement, ‘In Paris the PS resisted [La République en Marche], by uniting. So did the right, but by dispersing…’
The table above shows the right and centre-right lists in blue, but the results are not at all what you would expect. The official LR list was headed by Pierre Charon, one of the more colourful figures on the right and in the Senate, and an old member of Sarkozy's garde rapprochée. In 2011, Charon ran his own dissident UMP list. Now, it was poacher turned gamekeeper, as he faced not one but two LR dissident lists. The first, the Liste Républicaine et du Centre (RDC in the table above), headed by outgoing senator Philippe Dominati and Céline Boulay-Esperonnier, a veteran of the old RPR and UMP and a city councillor since 2001, actually outdistanced Charon and took two seats. The second, headed by Catherine Dumas (Liste Parisienne des Républicaines de la Droite et du Centre - PDC), took a fourth for the party. All four will probably join the LR group, but whether they are macroncompatibles or not remains to be seen. It is worth noting that the old Centre Party, the UDI, won no seats.
The real surprise package in Paris came from the Socialists (Liste du Parti Socialiste de Paris), who managed to bury their differences over Macron and, with the support of city mayor Anne Hidalgo, hold onto four seats. The Communists (through their secretary Pierre Laurent) and the Ecologists also held a seat each. In 2011, the left managed a broad, rainbow coalition list, but that was out of the question this time. And finally, Bernard Jomier, a former Ecologist, took a seat on a progressive ticket. It's not yet clear which group he intends to join when the lists are submitted later this week.
One final comment on the nature of Senate elections. It is not always the case, but often electors will try to balance up the representation of a department in geographical terms. Thus, in a department using the majority system, a Republican candidate for one district may well be elected with, say a moderate Socialist from another. This is less straightforward to achieve with the PR system, but usually, parties themselves will try to reflect the distinct indentities within a department by putting up candiates to appeal to these sub-constituencies. Sometimes though, it can go very wrong. Eight of the 11 senators elected this year for the Nord are from Lille.
So, why does this matter? Well, I mentioned in a previous post that the outcome of the renouvellement is, in a large part, the consequence of the Macron-Philippe administration's atttiude to local government and a determination to make huge budget cuts there. The right and centre-right have been very vocal in their opposition and their victory in the Senate election has given local government aassociations and élus a mandate to stand up to the government. last Thursday, at a meeting of the Conférence Nationale des Territoires, set up by Macron in the summer, PM Edouard Philippe received an icy reception when he announced that he stood by the decision to cut the grant to local government by 450 million euros. In response, the members of the council of regional presidents, Régions de France, announced that they would leave the conference, though the decision was not unanimous and has since led to its chair, Philippe Richert, resigning.
And while the election has been a boost for both Les Républicains and the Union des Démocrates et Indépendants (UDI), it has not been without collateral damage. On Sunday 1 October, in an interview with Le Journal de Dimanche, the UDI president, Jean-Christophe Lagarde announced that his party no longer considered themselves to be bound by any past agreement to Les Républicains. The Senate electoral campaign helped to underline the differences between the former allies. And the campaign for the leadership of LR will only serve to accentuate the gap...
(1) - Paul Smith, 'Et ce qui devait arriver arriva: The Senate falls to the left', in Sheila Perry and Paul Smith, 'Vivement dimanche: Media, Politics and France's Electoral year 2011-2012', Nottingham French Studies, vol. 52, no. 2, Summer 2013, pp.144-54.
(2) Paul Smith, The Senate of the Fifth French Republic, (Palgrave Macmillan 2009). See especially chapters 5 and 6.