Mini-Guide to the 2012 French legislative elections

The first round of legislative elections will be held in France on June 10, 2012; with a second round being held on June 17, 2012. All 577 seats in the French National Assembly (Assemblée nationale), the lower house of France’s Parliament, are up for reelection. These elections directly follow the presidential elections held on April 22 and May 6. Since the electoral calendar was shuffled up in 2000, the legislative elections have lost much of their importance and attractiveness as they are perceived as being mere confirmations by voters of their presidential vote a few weeks earlier. However, in constitutional terms, legislative elections remain extremely important for the President and his party, because, in the absence of a working majority for the president’s party, he is compelled to a cohabitation with the opposition, where the legislative majority and government is formed by the opposition.

Electoral system

The French National Assembly is composed of 577 members elected for five-year terms in 577 single-member constituencies. Somewhat at odds with the FPTP custom for legislative elections in countries with single-member constituencies, France has long used a modified form of runoff voting for legislative elections.

In each constituency, a candidate must win over 50% of valid votes cast (these are called suffrages exprimés) and over 25% of ‘potential votes’ or total registered voters (in French, we say 25% des inscrits) to win by the first round of voting. This means that is possible for a candidate to win over 50% of votes cast but not be deemed elected because he/she has not won over 25% of all potential votes due to high abstention. In 2007, an abnormally high number of deputies (110) won by the first round. Usually, only 10-20 deputies win by the first round.

In the event that no candidate has been elected by the first round, a runoff is held a week later opposing all candidates who won over 12.5% of registered voters (potential votes), or, in the case that only one or no candidate has won over 12.5% of registered voters, the top two candidates. In the second round, the candidate winning a plurality of the votes is elected. In case of a perfect tie, the oldest candidate is elected. Traditionally, runoffs usually oppose the top two contenders. The rising abstention in legislative elections, reaching 40% in 2007, means that it is increasingly hard for over two candidates to win over 12.5% of all registered voters. However, triangulaires opposing three candidates are quite possible. In 2007, there was only one triangulaire, largely because the far-right National Front (FN), usually the third party which partook in most triangulaires in the past, was crushed at the polls. In 2002, there were 10 triangulaires but in 1997, 79 of runoffs were triangulaires. It is possible and often quite common for potential triangulaires or even duels (normal runoffs) to feature either two or one candidate. This happens when dissident or allied candidates from either left or right qualify for the runoff, but choose to drop out of the runoff in favour of a stronger candidate from their political family.

This 12.5% threshold has been used since 1978. Since the 1978 elections, there have been no quadrangulaires (runoffs opposing four candidates). In 1958, when the legal threshold was only 5% of votes cast, only 20% of runoffs opposed two candidates – the rest involved three or more (up to six!) candidates.

It is important to point out that candidates in these legislative elections form a duo – there is the official candidate, but each candidate also has a suppléant (or substitute). A suppléant is a kind of alternate, elected alongside the candidate (in a sort of ‘ticket’), who will take up the deputy’s seat in the event that the deputy dies in office, is named to government (a member of cabinet cannot simultaneously be a lawmaker) or receives a public mission lasting over six months. In the case that a deputy resigns from office, for whatever reason, a by-election is held.

As per the constitution and Gaullist ideological tradition, the deputy is said to represent not his constituency or geographic region but rather “the nation” as a whole. In this vein, it is legal for a candidate to run for office in a constituency where he or she is not a registered voter. Voters might respond unfavourably to carpetbaggers, but no law prevents carpetbagging.

Redistricting remains a partisan government’s prerogative. Unlike in the United States, where redistricting is often done in a similarly partisan fashion, there is no law requiring redistricting to be done after a set amount of time or under a fixed schedule. Unlike in Canada or the United Kingdom, there is no independent electoral commission responsible for drawing constituencies. It is the prerogative of a political official – the Minister of the Interior – and of the government in power. The government decides whether there should be a redistricting or not. Since the single-member constituencies were first drawn up in 1958, there have been two rounds of redistricting: a nearly total revamp of the map in 1986, under right-wing Interior Minister Charles Pasqua, and a smaller but still significant redistricting in 2009, under Sarkozy’s then-secretary of state for territorial collectivities, Alain Marleix, who is, incidentally, the ‘electoral expert’ of the UMP. Hence, the map is a political creation of right-wing governments, who have not shied away from blatant gerrymandering to shore up their strongholds or limit opposition gains. The left can complain on some matters, but considering that no left-wing government has had the willpower to redistrict on its own terms, it should not complain too much.

Pasqua and Marleix’s redistrictings have been decried for the blatant gerrymandering. Not all departments are gerrymandered, but a good number of departments have constituencies which either poorly reflect geographic, social, economic and historical local realities or are blatant partisan gerrymanders. “Work of art” constituencies which are so common in the US are not very common in France, but Marleix drew a few ‘nice’ ones too in 2009.

A novelty in the 2009 Marleix redistricting was the creation of eleven constituencies for French citizens living abroad. In the past, French citizens living abroad could only vote in legislative elections if they maintained legal residence in French territory. The government opted to create eleven seats which would be elected directly by French citizens living abroad (who already have indirectly elected senators). There was a clear partisan motive behind this: French citizens living abroad are known for being (slightly) to the right of their compatriots in French territory. The right hoped that it could conquer about eight of the eleven new seats, crucial in the case of a tied game in France. In these elections, French citizens living abroad had the option of voting by internet – around 14% of registered voters did so. The election in these eleven seats plus the three constituencies in French Polynesia took place a week ahead of the rest: on June 2 or 3.

Context and Campaign

The presidential election on April 22 and May 6 resulted in the victory of the Socialist Party (PS)’s François Hollande, who defeated incumbent right-wing President Nicolas Sarkozy. The left’s victory ended ten years of executive and legislative dominance of the right. However, Hollande and his presidential majority need to win the legislative elections – and control of the National Assembly – in order to “wrap up” the victory on May 6 and control an absolute majority in the legislature. A parliamentary majority is crucial for any president who wants to have the means to pass key pieces of legislation.

As per custom, the President-elected named a Prime Minister and government who, despite lacking the confidence of a parliamentary majority per se, act as the full government of the country in the timeframe between the election of the President and the new legislature. Hollande named Jean-Marc Ayrault, a close ally and former parliamentary leader of the PS in the National Assembly, as Prime Minister. Hollande seemingly preferred the close ally, also a calm, reserved and competent ‘bureaucrat’, to a rival like Martine Aubry, the first-secretary of the PS. Ayrault’s government was certainly meant to reassure financial markets and moderate voters. The choice of a moderate social democrat, Pierre Moscovici, as Finance Minister, and a budget hawk, Jérôme Cahuzac as junior minister for the budget, reassured many. This was, after all, in part an ‘electoral’ government whose composition and actions would have some impact on the campaign and result for the legislature.

Hollande and his government have erred safe in this second electoral period, keeping away from the touchy subjects (like the budget and future spending cuts which will be needed) and going to great lengths to not alienate anybody while motivating the left-wing base. The government is aware that it cannot really begin to govern, so its first acts of governance have been largely symbolic measures like a 30% reduction in the salaries of the President and his cabinet or photo-op type sorties like those made by the flamboyant “minister for productive recovery” (the French left loves to give fabulous names to cabinet departments!) Arnaud Montebourg.

The inversion of the electoral calendar, established in 2000, set in stone the political primacy of the presidential election while reducing the importance of the legislative election. Indeed, being held only a few short weeks after the presidential election, during the honeymoon period, the legislative elections became confirmations of the presidential election result. In 2002, fresh from Chirac’s victory, the newly united right – the UMP – won a very ample parliamentary majority, ousting the incumbent left-wing majority which had just suffered an historic defeat in the presidential elections. In 2007, on the heels of Sarkozy’s comfortable victory, the UMP retained a large majority though the second round saw a very strong left-wing resistance and even offense. The inversion of the electoral calendar, by setting in stone the primacy of the presidential election, reinforced executive authority and presidential powers, in part at the expense of the legislature and the government which is responsible to it. Its rationale for doing so was to render extremely unlikely any future cohabitations, by assuming that voters would use the legislative elections to confirm their verdict in the big contest a month earlier. Since 2002, they have proven their political leaders. While anything is possible in politics, it would be the epitome of voters’ irrationality if they elected a new president but, a month later, turned around and denied him a majority. Of course, turnout differentials from one election to another renders it less totally illogical, but it remains that, when a majority of voters have chosen the representative of one political family over another, it is not very rational or logical for them to elect a government from the opposite political family.

The Parties and Hot Seats

The new presidential majority goes into these elections as the favourite. Hollande’s victory was narrower than expected and this underperformance will have its effect. It might even have its effect on the legislative elections. But for the time being, the government has a mini-honeymoon (though not a very festive one), and the general mood is one of anticipation of a left-wing victory in the legislative elections. Saying that voters do not favour a cohabitation is not entirely correct; right-wing voters would still very much like to have a cohabitation. But centrist voters – those who voted for Bayrou (but not only them) – would prefer a left-wing victory, though not a blank cheque like that granted by voters to the PS in June 1981 after Mitterrand’s first election.

The right does not hold much illusion about its chances. Like in 1981, the mood is defensive at best and salvage-what-can-be-salvaged at worst. The narrative about the UMP has moved beyond June to the upcoming fight for the presidency of the party, later this year, and the announced showdown between incumbent secretary-general Jean-François Copé and outgoing Prime Minister François Fillon. There are extremely few left-wing held seats where the UMP is in a realistic position to win the seat, therefore its entire campaign has been defensive, defending even what are either traditional strongholds (Lozère) or seats held by high-profile incumbents (Copé, Morano, Courtial, Novelli, Joissains, Mariton, Rosso-Debord etc). The UMP averaged about 30-33% in national polls (which are almost entirely useless and are awful creations of a media incapable of understanding legislative elections), which is roughly what the PS averaged (albeit the PS narrowly led most polls).

However, the UMP and PS have an unequal relationship with the fronts to their right and left respectively.

The far-right National Front (FN), whose charismatic leader Marine Le Pen won a record high 17.9% on April 22, is certainly looking to the legislative elections as a chance to confirm its result. In June 2007, a month or so after Jean-Marie Le Pen won a disappointing 10.4% in the presidential election, the FN suffered from mass demobilization and its candidate won less than 5% on average – truly rock-bottom for the FN, used to 8-10% at minimum since 1984. The FN clearly hopes to to follow up on its success on April 22 with another success in the legislative elections. However, traditionally, legislative elections have tended to be unfavourable for the FN, which, being very much a family business, suffers from a lack of high-profile local candidates and strong local grassroots. There are a good number of voters who vote for the FN only in presidential elections – I guess that comes along with the type of party it is – but despite the FN’s low-key paper candidates and its lack of solid institutional roots outside all but a handful of areas, it has managed to retain, especially since 1997, a strong vote in legislative and local (cantonal) elections. That being said, in 2002, Jean-Marie Le Pen won 17.8% in the runoff against Chirac but in the June legislative election, FN candidates won only 11.3% of the votes. In recent years, with abstention in these elections reaching new highs, a part of the FN’s wider base usually tends to abstain.

The FN cannot realistically hope to win more than 2 or 3 even in the best of scenarios, but it can certainly play a kingmakers’ role. The FN’s secret strategic objective is likely the defeat of the UMP, which it hopes to participate in by qualifying its candidates for as many triangulaires as possible. In 1997, when the party won a record 14.9% of the vote, 76 FN candidates qualified for three-way runoffs against left-wing and right-wing candidates, and in about six in ten of these cases, these triangulaires proved fatal for the right, and certainly contributed to the left’s victory. Therein lays the main threat of the FN in these elections, especially for the UMP.

The FN has taken on the etiquette “Rassemblement Bleu Marine” for these elections, a cheap rebranding effort (and play on their leader’s name) which officially aims to be a wider alliance of all nationalists but is in practice FN candidates with the addition of a few sidekicks. The FN is targeting a handful of constituencies with some high-profile star candidates, though it is an uphill battle for any of these star candidates. Besides Marine Le Pen in the 11th constituency of the Pas-de-Calais (Hénin-Beaumont), other star candidates include the former Eurosceptic MEP Paul-Marie Coûteaux in the Haute-Marne (2nd), Marine’s campaign director Florian Philippot in Moselle (6th), well-known lawyer Gilbert Collard in the Gard (2nd), Marine’s leadership rival and FN MEP Bruno Gollnisch in the Var (3rd), the former FN-turned-UMP mayor of Nice Jacques Peyrat in the Alpes-Maritimes (1st, though he is only supported by the FN and not a candidate of the RBM) and Jean-Marie Le Pen’s 22-year old grand-daughter Marion Maréchal-Le Pen in the Vaucluse (3rd).

The FN’s aim would obviously be to win a single seat, and though it has a lot of targets and a handful of seats which it could potentially take, it is an uphill battle in practically every constituency. Very much disadvantaged by the electoral system, the FN won only a single seat in 1988 and in 1997. However, the FN’s secret objective, again, is probably the defeat of the UMP, hoping to cash in on the internal chaos a very bad result would create and slowly take up the role of main right-wing opposition (a goal which is much more difficult to attain, clearly). The mariniste FN is playing very much on the anti-system, neither-left-nor-right line rather than the older mégretiste line of technocratic alliances with the parliamentary right. It will have no reluctance in going out to defeat UMP incumbents, even if the left would be the sole beneficiary of that, because it brands the UMP and PS as one and the same in a homogeneous political establishment.

The FN’s nuisance power depends on the amount of loses it suffers to other parties vis-à-vis Marine’s April 22 performance, and also on the overall turnout level. The FN is currently polling 14 to 16%, but it is quite possible that it is suffering the heaviest loses compared to April 22 in the regions of the country where it is weak and where Marine was more likely to touch an ephemeral, non-partisan protest vote than an engrained traditionally frontiste vote. If this proved to be the case on June 10, the FN’s nuisance power would be fairly significant because the regions where it is most dangerous for the UMP are those regions where it is traditionally strong and well implanted. However, in order to stage the most three-way battles as possible, turnout must be fairly strong – at least over 60% (the 2007 level). In a 60% turnout hypothesis, the FN would need about 20% of votes cast in order to qualify for the runoff. If it polls only 14% nationally, it will be hard to attain this level in a large number of constituencies. 100 appears to be the maximum number of possible three-way runoffs with the FN, 50-70 appears like a reasonable estimate while 30 is likely the most conservative estimate for the number of three-way runoffs with the FN.

On the other hand, the other front – the Left Front (FG), whose fiery leader Jean-Luc Mélenchon won 11% in the presidential election – has a much calmer relationship with the PS. Even though the FG is running independent candidates in all but two metropolitan constituencies (after a potential deal on common PS-FG candidates in shaky constituencies collapsed), which has traditionally been the norm for the PCF, the unwritten tradition would call for the FG (or the PS, depending on the case) to drop out and/or endorse the first-placed left-wing candidate in each constituency. Mélenchon might play hardball way more than the PCF did in recent years, but when push comes to shove, he can be counted on being a good soldier and line up behind the wider presidential majority without too much trouble. Unlike the FN which, it believes, has something to gain from the defeat of the right; the FG certainly has nothing to gain at this point from a defeat of the left. FG candidates and their voters will line up behind the first-placed left-wing candidate in practically all cases, and triangulaires opposing a FG and presidential majority candidates are very unlikely. The FG is part of the wider plural left, the FN cannot even be counted as part of the wider parliamentary right with the UMP and its smaller allies.

Thus, while the UMP cannot count on a smooth transfer of FN vote even in traditional runoffs and must fear the prospect of three-way battles, the PS and its close allies can count on the smooth transfer of a vast majority of FG votes, quantified at 6-9% in most national polling, in almost all constituencies.

Jean-Luc Mélenchon became known during the presidential campaign for his frontal opposition to the FN and Marine Le Pen, and the two have developed mutual hatred (this is no exaggeration) of one another. In part to annoy her, in part to keep himself in the spotlight and in part to actually win, Mélenchon himself opted to run in the 11th constituency of the Pas-de-Calais, against Marine Le Pen in her own backyard. Though Mélenchon has no roots in the area, he aims to benefit from the local divisions of the PS in Hénin-Beaumont and the department in general. Indeed, despite the weak implantation of the PCF in this area, he could hardly have chosen a better place to land in. The local PS has been in shambles since 2009, and the PS nomination fight in the constituency was marked by intense divisions, personal feuds and much recriminations. The PS candidate, Philippe Kemel, won a contested primary, but his support from the old leadership and local base is limited at best. The retiring incumbent’s silence likely amounts to backhanded support for Mélenchon and the DVG mayor of Hénin-Beaumont is almost public in his support for Mélenchon. The bitter and nasty contest in the 11th constituency has been nationalized by the candidacy of the two fronts‘ respective leaders.

Outside Mélenchon’s race, the FG hopes to pick up a few seats in departments such as Seine-Maritime (6th), Ardèche (3rd), Bouches-du-Rhône (7th, 10th), Pas-de-Calais (7th), Nord (19th) or Jura (3rd).

The FG’s aim in this election is to win a significant number of seats (at least 20) in order to weigh more heavily on the wider parliamentary left. A 1988-like scenario where the PS, the Greens and its other allies and dissidents hold only a minority of seats (but a plurality over the right) would wet the FG’s appetite as they would be necessary legislative partners for the PS (especially with the upcoming annihilation of the MoDem) and could throw their weight around with much more impact than if the PS and its allies won an absolute majority on their own.

The FG will find itself divided after the elections over the question of potential cabinet participation. The present cabinet, besides two Greens and two Left Radicals, is a unicolour PS government. The PCF, whose leadership has since 1997 been keen on the plural left, is open to cabinet participation under certain conditions. On the other hand, Mélenchon and the PG are very hostile to any cabinet participation and much prefer to play the role of picky legislative ally for a PS (minority) government that the PCF played between 1988 and 1993 with the PS.

The PS’ goal is obviously to win an absolute majority alongside its closest allies, the PRG, the miscellaneous and dissident lefties and the Greens (EELV). This goal is quite attainable.

However, the PS has been marked by a very large number of dissidences throughout the country. The root of the problem is the November 2011 deal with EELV, in which the PS conceded (meaning that the PS officially backed a EELV candidate) about 63 constituencies, with 20 of them winnable, to EELV. This deal was often done over the heads of local PS federations and caused much controversy on the ground, given that a lot of the EELV candidates in these conceded constituencies had little to no local political implantation, while the local PS often had fairly strong benches of local mayors, general councillors and so forth. The fact that EELV gave almost nothing away in return – running candidates against official PS candidates in other seats – caused more controversy.

As a result, almost all constituencies given to EELV in the November deal have one or more PS dissident candidates, candidates who were promptly excluded from the party by the national leadership. The most high-profile dissidences include the 1st constituency of the Rhône (where the PRG’s Thierry Braillard, backed by Lyon PS mayor Gérard Collomb, is running against a EELV candidate backed by the PS), the Côtes-d’Armor (4th) and the Nord (8th, where the PS incumbent is running as a dissident against a EELV candidate endorsed by the PS). In Paris’ 6th constituency, the PS incumbent who got shafted by the deal ultimately bowed down to pressure and accepted to become the suppléant of the EELV candidate… Cécile Duflot, the party’s leader and cabinet minister (in the end, Duflot will win but her PS suppléant will ‘hold’ the seat…). The dissidents have a strong argument on their side: the deal in November was signed when EELV was polling 5-6%. On April 22, EELV won less than 2% of the vote.

EELV will still hope to get at least 15 seats this election, the minimum required to form a parliamentary group. But with the presence of stronger PS dissidents in almost all constituencies, it could end up winning as few as 10 seats. Still, it is hard to say whether voters will vote in a “legitimist” fashion and backed the EELV candidate who is officially backed by the PS (and this mention likely features prominently on their ballots…), or if they will prefer voting for a well-known locally based dissident. If EELV fails to win a parliamentary group even after a generous deal with the PS, it will be thrown back into its old state of utter dependence on the PS, making the 2009-2011 episode seem like a distant wet dream.

The PS dissidents, often way better locally implanted than the official EELV-PS candidate, will probably far outrun the EELV candidate. In the end, however, as always, the PS dissidents will soon find their way back to their native party and be active members of the presidential majority. There are other cases where a strong local PS dissident is running against a PS canddiate. Most notably, there is Olivier Falorni in La Rochelle, a local dissident against Ségolène Royal, the 2007 presidential candidate and open candidate for the presidency of the National Assembly, who was endorsed by the PS in the solidly left-wing constituency of La Rochelle, despite being from the neighboring Deux-Sèvres. The other main PS vs. PS fight is in the Aude (2nd), where the frêchiste Didier Codorniou, excluded from the PS since 2010, is running as a dissident against a candidate officially backed by the PS.

The traditional deal with the PRG created much less animosity. The PRG and PS have long been very close allies, to the point where the PRG has basically turned into an annex of the PS’ “right-wing”. The deal between the two gave 32 seats to the PRG, 20 of them deemed winnable. Few of these deals gave rise to major dissident candidacies, because the PRG’s candidates are often well implanted local officials (former deputies, local mayors, general councillors etc) unlike the majority of the EELV candidates backed by the PS (most of whom hold no other elective office). With this deal, which can secure an additional 5-10 seats for the PRG (which won 7 in 2007) the PRG can hope to form a parliamentary group on its own or with the possible addition of a few DVG and PS dissident candidates. Indeed, the PRG has often acted as a receptor for some wandering non-socialist leftists or unhappy ex-socialists.

The centre, as always since 2007, is a disunited mish-mash. François Bayrou’s MoDem is running candidates in a vast majority of constituencies, but under a new empty label, the Centre pour la France, which, like the FN’s RBM, is largely another label for what are, in large majority, MoDem candidates. However, the MoDem can hardly be playing offensive at this point in time (besides for a few strong high-profile candidates like Rodolphe Thomas or Gilles Artigues who are backed by the UMP in their respective PS-held constituencies) when their own leader, Bayrou, who personally endorsed Hollande between the two rounds, is facing his toughest race in years. Bayrou’s decision to back Hollande, like every other political decision he has made since 2007, has backfired badly on him. The UMP is angry with him, while the PS has taken to ignoring him by running a candidate against him in his constituency despite his endorsement of Hollande. Polls have shown that he is either neck-and-neck or narrowly trailing this PS candidate in his Pyrénées-Atlantiques (2nd). Bayrou had won only 19.9% and third place in his constituency on April 22, when he had taken over 39% in his home turf in the 2007 presidential election. His neighbor, the flamboyant Jean Lassalle, is also threatened by an aggressive local PS which fancies its chances of winning all constituencies in the increasingly pink Pyrénées-Atlantiques.

The MoDem’s candidates finds themselves marginalized in these elections, which have long been polarized betweeen left and right. A large number of Bayrou’s voters from April 22 prefer to vote either for the UMP or the PS, while the MoDem struggles to find a voice in this contest, not sure which way to play. It hesitates between the old neither-left-nor-right centrism which hasn’t worked and the vague indications of centre-leftist leanings which won’t work. Bayrou’s messy calls for some type of “constructive” force which is neither in the opposition nor in the majority have obviously been squashed in a polarized election.

On the centre-right, the NC has been compelled to its old state of unequal dependence on the UMP. All but two of its 20 something incumbents are unopposed by the UMP. While the situation is not as straightforward in all constituencies, for all intents and practices, the NC in this election is hardly distinguishable from its old big brother, the UMP. The NC is running just enough candidates of its own to save its public financing – a really big deal which is based on votes cast for your party in the legislative elections, but the NC will come out of these elections in its same state of subjection to the UMP’s diktat. With a sizable number of NC deputies in close races, there is a good chance that the party will emerge only marginalized and weakened from the elections. The fact that the NC’s parliamentarians (which are basically the only NC members of renown) will have been reelected or elected only thanks to the UMP’s support by the first round, will continue to complicate efforts at centrist re-foundation because NC deputies will be wary of cutting bridges with the party which allowed them to hold their seats!

Jean-Louis Borloo’s Radicals (PRV) are in a similar state, though showing more independence against the UMP. Though the PRV’s incumbents (including Borloo, himself in a marginal constituency in Valenciennes) are, in large part, unopposed by UMP candidates (a fact which explains why Borloo’s attempt at Radical independence were coolly received by a lot of PRV parliamentarians…); in a few constituencies it is running candidates against the UMP, candidates which are sometime clinging to the stillborn ‘ARES’ label. The most high-profile of these cases is in the Hauts-de-Seine (2nd) where Rama Yade, the former cabinet minister turned Borloo ally, is running against UMP incumbent Manuel Aeschlimann in a seat which will likely be won by the PS’ Sébastien Pietrasanta.

Legislative elections are big deals for all parties, because they decide the public funds each party receives. A party needs to run a set amount of candidates and receive a not-insignificant amount of votes to be eligible for public financing, so practically every party – ranging from the big ones to tiny ones you would never have known about – are running some candidates. In all but a few cases, only the major and ‘major minor’ parties will win seats or be remotely competitive in each constituency (in metropolitan France).


I don’t often – if ever – do predictions on this site, largely by fear of being humiliated at how bad my predictions were. Pushed by the incompetence of the media in covering these elections, I’ve decided to take a big risk this time and build my own predictions of the final winner, constituency by constituency. Before the first round, to err safe, I have only done predictions on the basis of “political families” (left, right, centre, far-right) rather than by political party. I have examined each constituency and built a prediction based on qualitative personal factors such as local candidates, grassroots strengths of the various candidates and any personal knowledge of politics in the constituency; quantitative political factors such as the impact of redistricting (if any), the results of the presidential election (first and second round) and the results of the 1997 and 2007 legislative elections (if applicable). I am quite partial to the American way of classifying constituencies as “safe”, “favoured”, “lean” or “tossup” and have used those classifications here.

Safe seats for the left or right are seats where the chances of another political family (like the right in a left-held seat) of winning the seat are extremely low, nil or downright impossible. ‘Favoured’ seats indicates seats where a political family is ‘favoured’ to win the seat, and where it is quite or very unlikely that another political family will win the seat. ‘Lean’ seats indicates seats where one family has an edge or significant advantage over another political family, though it is not unfathomable to see the other side winning.

I came up with two classification of “tossup” seats, which could go any way without much swing. There are ‘edge’ tossups where, though the seat remains very disputed, a political family (the left, right, centre and so forth) has a narrow or consistent advantage, a marginally better chance of winning in the end. Pure tossups are those few seats where it is very much impossible for me to decide on a side which holds even a marginal edge, even less a final winner. Some pure tossups are three-way races including the far-right.

For French Polynesia and French citizens abroad, the map and numbers depicts my final prediction based on first round results.

My predictions give the following numbers, as of today, for the wider political families:

Safe left: 210
Left favoured: 42
Lean left: 30
Tossup – left edge: 43
Parliamentary Left (FG+PS+DVG+PRG+MRC+EELV): 325

Left-right pure tossups: 12
Left-right-far right pure tossups: 3
Right-far right pure tossups: 2
Pure tossups: 16

Tossup – centre edge: 2
Centre-MoDem: 2

Safe right: 85
Right favoured: 67
Lean right: 41
Tossup – right edge: 40
Parliamentary Right (AC+NC+PRV+UMP+DVD+DLR): 233

My predictions would give about 55% of the seats to the left overall, with an outside chance that the PS, DVG, PRC and EELV could hold an absolute majority (289 seats) on their own without the need for the FG’s votes.

You will notice the disparity in safe seats between left and right. I have simply been very reluctant to classify seats currently held by the left as anything other than “safe” in this climate, and similarly reluctant to grade too many seats held by the right as being completely “safe”. In seats where the right is only favoured, I feel as if there is a very/extremely remote chance for there to be surprises or even an upset, though I certainly do not expect any. We can never be too safe, especially when we do not yet know (for sure at least) whether the left’s likely victory will be a tsunami or only a normal little tide.

Based on my predictions, all Ayrault cabinet ministers who are candidates in these elections would be reelected – including Marie-Arlette Carlotti who is running against a UMP incumbent in Marseille, who is certainly the cabinet minister with the toughest fight. There is a chance that junior minister Kader Arif could be defeated, but by fellow left-wing dissidents rather than by the right – he is running in the Haute-Garonne. Stéphane Le Foll, the new agriculture minister, should conquer François Fillon’s old seat in the Sarthe.

According to my predictions, I feel as if Frédéric Lefebvre, Chantal Brunel, Guillaume Peltier, Valérie Rosso-Debord, Hervé Novelli, Eric Raoult, Georges Tron, Jean-Christophe Lagarde, Laurent Hénart, André Santini, Hervé Mariton and Maire-Anne Montchamp would be defeated. I still give a narrow edge to other contenders in very tough races, including Nadine Morano, Christian Vanneste, Yves Jégo, David Douillet, Jean-Louis Borloo, NKM, Xavier Bertrand, Laurent Wauquiez, Michèle Alliot-Marie and Patrick Devedjian. François Bayrou’s seat is rated as a “pure tossup”, not a very dangerous call…

I do not predict any far-right winner, though I feel as there are a maximum of five seats where the FN could realistically win. However, I have classified the 11th constituency of the Pas-de-Calais, where Marine is running, as a “tossup – left edge”, with Mélenchon the likely winner. I think that Jacques Bompard, the ex-FN mayor of Orange, who is not actually a FN candidate, is the far-right candidate with the biggest chance of winning (in Vaucluse’s 4th), with Paul-Marie Coûteaux standing a fairly good chance in a “pure tossup” race in Haute-Marne’s 2nd.

Besides Mélenchon in the Pas-de-Calais, some other “big winners” based on my predictions include Jack Lang (in the Vosges now…), Patrick Mennucci, François Fillon (in Paris’ 2nd instead of Sarthe this year), Cécile Duflot (in Paris’ 6th) and potentially Ségolène Royal.

My predictions remain, of course, subjective and based only on limited knowledge. I am open to justifying my calls in all constituencies, explaining my predictions further or receiving comments and feedback based on my calls in such and such constituency.

A big question mark ahead of June 10, and a question mark which has a very big impact on these predictions, is the issue of triangulaires with the FN. Nobody knows how many there will be, and estimates range from lows of 30 to highs of 100-110. It is quite hard to predict how many there will be, it is even harder to predict with much certainty where they will be. At any rate, 100-110 seems high (though more reasonable than the media’s sensational headlines on April 23 about ‘over 300 triangulaires’!), considering that in 1997, with 14.9% nationally and 67.9% turnout (which will not happens this year), the FN qualified for 76 three-way runoffs.

For my analysis, I have worked on the assumption of a national FN result around 15% and 60-63% turnout, and used Marine’s first round results as the basis everywhere. Of course, this is far from a perfect strategy to approach this problem. UNS is a myth which doesn’t exist and will never exist, so assuming that you can just subtract 3-4% from Marine’s vote everywhere and get a solid prediction of the FN vote on June 10 is pretty stupid. If the FN vote drops the most in constituencies where the FN has not traditionally been very strong but where Marine was able to win some “new” voters, but stays at solid April 22 levels in core FN territory, then the FN will be in a good position to qualify for many triangulaires.

The second issue, which will need to be examined in the wake of first round results, is how the FN will perform on June 17 in those constituencies where it is qualified for either duels or triangulaires. In the past, especially in 1997, the FN had tended to lose a fairly significant percentage of its first round vote in triangulaire runoffs (I’m not aware of a statistical study on this phenomenon, but in 1997 it was almost universal in all 76 FN triangulaires).

There was a tendency for a sizable minority of FN voters to vote for the FN in the first round but to either sit out the runoff or switch their vote to the left or right-wing candidate who has a stronger chance of winning. FN voters are not necessarily locked into a “all the others suck” mindset themselves, and some use the FN as a first round protest vote, but pragmatically support an electable candidate from either left or right in the runoff. Just because their party is isolated behind an old cordon sanitaire doesn’t mean that FN voters don’t conform to the old rule of “in the first round, you choose; in the second round, you eliminate”… even when they could still “choose” rather than “eliminate” in triangulaires.

In the 2010 regional elections, however, the FN gained votes compared to the first round in (I think) every region where it qualified for the runoff (all were, obviously, triangulaires). There were only very few triangulaires in the 2011 cantonals, so it is hard to see if this trend was ephemeral or if it is going to stick. If on June 17, the FN tends to lose votes compared to the first round in the triangulaires where it is qualified, the UMP candidates stuck in these triangulaires de la mort could hold out hope that some first round FN voters will pragmatically vote for the UMP over the FN in a runoff situation favourable to the left.

We will also need to look at how the FN performs, compared to the first round, in those constituencies where it is alone against either the left or right in the runoff. The myth is that left-FN runoffs are way more favourable to the FN than right-FN runoffs are. Besides the fact that the two cantons the FN won in 2011 were held by the left, there is little proof that this is actually true. In the 2011 cantonals, the FN gained an average of 10.6 percentage points in left-FN runoffs, while in right-FN runoffs, the FN gained 10.5 percentage points on average.

The theory that mainstream right voters are more likely to vote for the FN in left-FN runoffs than left-wingers are in right-FN runoff makes a good deal of sense, but I’m not sure if it has been proven quantitatively. The 2011 cantonal elections and the 2012 presidential elections showed that the FN attracts a very ideologically, socially and geographically diverse electorate. The 2011 (and even 2004) cantonal elections further showed that the FN was an attractive option for a sizable minority of left-wing voters in right-FN runoffs. This, plus the interesting trends noted in 2010 triangulaires is part of wider phenomenon of destigmatization of a FN vote, what is described in France as a lepénisation des esprits (a ‘Lepenization’ of mainstream political thought). The FN, for a good deal of voters, is not as repulsive as it was in the 1980s and 1990s, and a lot of voters might (secretly) agree with some of the FN’s rhetoric and ideology, given the general conservatism of the wider electorate on law-and-order and immigration issues. For opponents of the FN and its politics, this is a very worrying trend, but as far as we are concerned, it explains why the FN has tended, since 2007, to perform increasingly well in runoff scenarios, including triangulaires. 

The results in the 11 new “foreign” seats have fallen, and they are very – surprisingly – favourable for the left. These elections, which included an option of voting through the internet, were marked by extremely low turnout – about 15-25% of registered voters turned out, a majority voting online. When turnout is this low, you open the door to weird results and very unpredictable elections. It appears as if demobilization was heavy on both sides, but right-wingers in these eleven ‘foreign’ seats were more demobilized than left-wingers. It is a logical pattern, which is nothing new (it happened in 1981), but it does not seem to have been picked up by pollsters in France, which do not report spectacular levels of right-wing demobilization. However, if these patterns of mobilization hold up in France on June 10, the PS and its allies can hope for a large victory – unless the runoff “corrects” the first round, like in 1967 or 2007.

In these seats, it appears as if the UMP’s tactic of ensuring itself at least 7-8 seats will backfire badly. Based on first round results and personal analysis, I predict that only three of the eleven constituencies will return UMP members – the rest will return left-wing members. The UMP suffered a lot from poor candidate selection, its preference of ‘metropolitan’ candidates perceived as carpetbaggers over locals (who often ran as dissidents) being a particularly poor strategic choice. This was the case in the first constituency, covering Canada and the United States. The UMP candidate, Frédéric Lefebvre, a Sarkozyst and former cabinet minister, found himself very contested on the right by high-profile dissident candidacies. In the end, Lefebvre, never a good candidate for anything, won 22.1% and distant second behind the PS candidate, Corinne Narassiguin, a strong local candidate, who won 39.7% in a seat usually favourable to the right. She is the likely winner in the runoff.

Another former cabinet minister running for the UMP, Marie-Anne Montchamp, will likely suffer a similar fate in another constituency which the right should have won easily (the Benelux, 4th). She won only 21.2%, distant second to the PS who won 30.4%. With transfers from the Greens, the PS will likely win this seat, again a seat which the UMP didn’t really have any business in losing. On the topic of this same constituency, Dominique Paillé, a former UMP deputy (defeated in 2007) who has since sided with Borloo’s PRV-ARES, ran in this seat and won… 2.79%. While Dodo has lost every election he has run in since 2007, this is a new low for him.

Most surprisingly, however, is the eight constituency, which covers Italy, Greece and Israel. Backed by over 90% in Israel, this constituency was Sarkozy’s best constituency out of these eleven new seats. Yet, due to very low turnout in Israel, the PS won first place with 30.5% against 22.5% for the UMP, in a very spot for the runoff. A likely left-wing victory in this constituency would be a perfect symbol for the right’s ghastly performance in these eleven new seats. Even stronger high-profile candidates for the UMP like former judge Alain Marsaud and former cabinet minister Thierry Mariani are not fully certain of winning their own seats.

For those looking for a guide of seats to follow when results flow in on June 10, I have prepared this little colourful guide of constituencies which will be of some interest on June 10 (omitting some, naturally).


Posted on June 7, 2012, in Election Preview, France. Bookmark the permalink. 11 Comments.

  1. Thanks a lot for the post. Was worth the long wait

  2. Great post as always, thanks. I’m no party hack (or even a French voter) but I was talking to Jean-Christophe Lagarde’s campaign manager a few weeks ago and when I asked him about his candidate’s chance of success he was supremely confident of victory. He could have been bluffing but maybe there’s a strong local factor in that constituency that’s going to buck the national trend.

  3. Georg Ebner delivers similar projections.

  4. You write:

    “Hence, the map is a political creation of right-wing governments, who have not shied away from blatant gerrymandering to shore up their strongholds or limit opposition gains. … Pasqua and Marleix’s redistrictings have been decried for the blatant gerrymandering. Not all departments are gerrymandered, but a good number of departments have constituencies which either poorly reflect geographic, social, economic and historical local realities or are blatant partisan gerrymanders. “Work of art” constituencies which are so common in the US are not very common in France, but Marleix drew a few ‘nice’ ones too in 2009.”

    Can you give some examples?

  5. @Bancki:
    Somme-1 and 4 are truly the most horrible ones. Vaucluse-5 is awful. Gard-1 and 6 are clear gerrymanders. Pyrenees-Orientales 3 doesn’t look bad but makes no sense. You also have some less aesthetically ugly ones, but which form part of Marleix’s “packing” strategy – concentrating left-wing strongholds in one constituency: Herault-2, Bouches-du-Rhone 4, 7 and 13, Ille-et-Vilaine 2, Seine-Maritime 5 and 8, Nord-8 and 13. The way the map looks like around Orleans, Metz, Laval, Bourg-en-Bresse, Limoges, Toulouse and Rennes is quite bad: the old way of dividing major cities into two or more constituencies with the addition of suburban/rural areas.

  6. I’m interested in some of the reasons why the Socialists have never redistricted themselves. Is it part of a holdover of the left’s long dislike of the Fifth Republic structures, that changing the boundaries would throw open the question of changing to PR (as Mitterrand did before the 1986 election?)

  7. As I rely on the opinion polls to understand the public mood and the potential outcome of the election, I’m wondering why you state that you have no confidence in these? I know that various opinion polls can sometimes by out by a significant margin, but are French pollsters unaware of how to conduct ‘scientific’ opinion polling, which should closely reflect public voting intention? Are there any French pollsters which are reliable, and more weight be placed on their polling data than that of others?

  8. @CTerry: The left could have redistricted twice since 1986: between 1988 and 1993 and between 1997 and 2002. Of course, in 1988-1993, the Pasqua redistricting was ‘fresh’ and it was never on the agenda and I think that’s fair enough because you won’t redistricting only 2-7 years after it was done. The left could and should have done a redistricting between 1997 and 2002, because it was over 10 years after the fact and the population distortions were getting important.

    Redistricting is not a major political issue in France: nobody except some bureaucrats and 5 people who like election maps care about it, and there is no law which structures the process so there’s no real obligation for anybody to do it.

    As for the left, I think the Plural Left structure of the Jospin government might have been a thorn in the side of a redistricting effort. Firstly, you would have some Socialists clammoring for gerrymandering some PS-winnable seats out of PCF strongholds; you’d have the PCF wanting to get more seats; while the Greens would have argued in favour of PR and could potentially throw a fit about a redistricting which doesn’t change the electoral system. On the other hand, Sarko’s “coalition” (or Chirac II in 1986) had no parties which favoured PR or were all that big on electoral reform unlike the Greenies are.

    Anyhow, as always, the left always complains very loudly about how atrocious and unfair the right’s electoral or institutional reforms are, but they never change anything. I doubt Hollande’s government will touch the electoral reform issue, especially if the PS+DVG+PRG have 289 seats or more. They might change Sarko’s ‘territorial reform’, but even then…

  9. @Mrodowicz: National polling is good for presidential and European elections, more useless for legislative and regional elections and totally retarded for municipal and cantonal elections. The best explanation for why I very much dislike national polling for legislative elections is that these electionS are basically “577 local elections on the same day”. Each constituency (less so today than in 1958, for sure, but still…) basically has its own election. Candidates of the main parties vary in strength from place to place, incumbents are stronger than their national party in some cases, some parties have no local infrastructures and can’t cash in on their “potential” votes, local factors play a major role, and of course party alignments are different from place to place.

    For example, EELV has 63 candidates backed by the PS, but in other constituencies EELV has no candidates. The NC polls 1% nationally, but it has all but 2 of its 25ish incumbents backed by the UMP. When pollsters do their polls, I’m pretty sure they ask a generic national ballot question. Certainly you’ll have many people who say they’d vote PS in a national ballot, but in their home constituency there is no PS candidate…

    Really, national polling here is only useful for the FN, whose vote is not all that dependent on local candidates (if it was, considering some of their candidates… they’d poll well in 5 constituencies and 1% everywhere else). As we’ll see when I draw up maps of the results, the UMP/PS/MoDem/centrist/EELV/PRG/PCF vote is really dependent on local factors such as incumbency or candidate strength. You’ll have the UMP candidate polling like 15-20% max in Mayenne’s 2nd constituency (which is a right-wing stronghold) but something like 40-50% next door in the 3rd – and not because one is more right-wing than the other.

    Of course, national polling has some sense in quickly seeing the tendencies and party strengths, but that’s all there is to it. It is still, however, less stupid than national polling for municipal or cantonal elections which is downright retarded.

    In terms of pollsters in general, my friend’s ratings based on the first round had Harris (!) in first place, IFOP in second, OpinionWay (despite lefties decrying it as the right-wing media’s pollster) in third. LH2, BVA and CSA were, unsurprisingly, the worst.

  10. I think the left didn’t touch the districting in ‘their’ 1997-2002 legislature, also because malapportionment (overrepresentation of districts with declining population numbers) was working in their favor, and the right did start a redistricting largely for the same reason (and for the 11 expat seats).

  1. Pingback: France (Legislative) 2012 « World Elections

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

%d bloggers like this: