Judicial impartiality in politically charged cases

On the relationship between judicial impartiality and court legitimacy, see, e.g., Gibson and Caldeira (2012).

It must be noted that the CE’s decisions are not subject to the sample selection issue raised by Priest and Klein (1984). They argue that cases which proceed to a judgment are not a random sample of disputes because disputants can settle their disagreements out of court in light of the expected litigated outcomes. However, it is impossible for winners and losers in mayoral elections to settle their disputes out of court. Therefore, the observed verdicts in this study are really based on all the disputes triggered by controverted elections.

On the organization and the judges of the CE, see Sect. 2.1 below.

There is a disagreement in the literature, reflected in the works of Berkowitz and Clay (2006) and La Porta et al. (2008) among others, as to whether judges are granted a higher level of independence from the executive and legislative branches of government under British common law than under French civil law. This issue is however beyond the scope of our study which focuses on the supreme administrative court of France and does not involve comparisons between different legal systems.

It must be noted that in France, it is the Supreme Constitutional Court—the Conseil Constitutionnel—which evaluates the validity of the presidential and parliamentary elections. See, e.g., Franck (2009, 2010) and Espinosa (2017)..

For instance, Régis Debray, whose background is in philosophy and not in law, was nominated in 1985 to the Conseil d’Etat. His close relationship with Danielle Mitterrand, wife of then-President Francois Mitterrand, was the obvious reason of his appointment. See Pfister (1986) on this.

Although there is no precise statistics, it is usually thought that roughly half of the current judges entered the CE through the “direct” way while the remaining half entered it through the “indirect” way.

It is the case that the other known instances of governmental interferences with the CE’s independence occurred in times of political upheavals and wars. In 1879, i.e., in the first years of the Third Republic, the Republicans who intended to get a hold on French institutions conducted an extensive purge of the CE to get rid of the Bonapartist and Royalist judges (Wright 1972). During World War II, the Vichy government which was allied with Nazi Germany dismissed Jewish judges (Baruch 1997).

As we discuss in an online Appendix (Appendix A), voters do not directly elect the mayor. Instead they vote for the members of the mayoral council who then elect the mayor and his/her deputies. Such a system is possible because candidates in mayoral elections compete on lists, with the head of the list being the would-be mayor of the town. Therefore, although it would be more precise (and more cumbersome) to write about "the heads of the lists competing in the elections to the mayoral council", we refer for simplicity throughout the paper to "the candidates running for mayor in the mayoral elections".

There are 41 TAs, at least one per région (which is the administrative division created in 1982 to subsume départements). The 31 December 1987 law instituted five Cours Administratives d’Appel (three additional Cours Administratives d’Appel were created in the following years), so that the CE would not have to judge all the appeals of the TAs’ rulings. However, in matters of controverted elections, the appeal of a TA ruling is still directly judged by the CE.

It must be noted that the CE mentions the TA’s ruling when it rules over the validity of a controverted election. However the CE does not actually take into account the TA's verdict when handing out its own decision because it reexamines both the facts surrounding the election and the laws which were supposedly breached.

Note that only some of the CE rulings are published in the Recueil des décisions du Conseil d’Etat, also known as Recueil Lebon. It is worth noting that our dataset comprises many CE rulings which were available in the CE archives but were not printed in the Recueil Lebon.

In an online appendix (Table SA2, Appendix C), we provide the numbers of invalid and valid requests between 1958 and 2007, the numbers of elections with irregularities and of cancelled elections, as well as the percentages of irregular and cancelled elections out of the total number of valid requests.

We cannot ascertain whether the candidates commit irregularities because they break electoral rules which they are not aware of, or because they are dishonest and cheat intentionally.

We cannot run tests on the actual rates of cheating of each party, because some candidates may cheat and win the elections without the losers filing a request on the validity of the electoral race. But given the tendency of losers to complain to the CE, it is reasonable to assume this rarely happens.

The t test of means comparison in Panel A of Table 4 also show that elections of mainstream right-wing politicians have a larger vote differential than those won by unaffiliated right-wing politicians, as well as by regionalist and far-right candidates, but the results are found not to be robust in the non-parametric test of means comparison shown in Panel B of Table 4.

Duverger (1986), among others, discusses how the constitutional arrangements of the Fifth Republic may entail a division of the polity between left-wing and right-wing parties.

While the President of the French Republic’s veto power does not allow him to unilaterally rescind legislation that he opposes, he still wields enough influence to partially disrupt the policies undertaken by his own Prime Minister. Indeed, the veto power of the French President is restricted by Article 10 of the French Constitution which states that the French President has to promulgate a law passed by the Parliament in the 15 days that follow the vote. In this time period, he can ask the Parliament to re-discuss the law or file a request to the French constitutional court, known as the Conseil Constitutionnel. But if the Parliament adopts the same bill or if the Conseil Constitutionnel rules that the law is in conformity with the constitution, the President has to comply and enact the law.

As mentioned in Sect. 2.3.1, the CE does not take into account the TA's ruling when handing out its own decision because it reexamines both the facts surrounding the election and the laws which were supposedly breached. This specific judicial procedure (as well as specification biases which arise within the Heckman model) explains why we do not include the TAs’ decisions as an explanatory variable in our empirical study.

The regression results reported in Columns 1 and 2 of Table 6 are not robust over the two samples (with and without the by-elections). They suggest that the CE judges did not reject the complaints of mainstream right-wing politicians on procedural grounds (in the whole sample) but that they dismissed on procedural grounds the requests against the elections of communist candidates (in the whole sample) and of left-wing politicians unaffiliated with a national left-wing party (in the restricted sample).

It is unlikely that the division of power between the government and the parliament would lead CE judges to be more diligent in the application of legal procedures.

The results are identical if we start the sample after the 30 July 1963 reform of the CE. This finding is not surprising since we noted in Sect. 2.2 that the 30 July 1963 reform was more technical than substantial.

In our specification, the variables pertaining to the authors of the requests are included in the first step of Heckman's two-step model but those related to the winners of the elections are included in both steps. Such a specification allows us to determine at which stage of the judicial review process the political affiliation of the winners influences the CE judges. Furthermore, this specification allows us to determine whether the ruling coalition benefits from the cancellation of the election. As such, in Columns (1), (2), (5) and (6) of Table 6we consider that the right-wing government benefits from the victories of mainstream right-wing candidates and of right-wing candidates who are unaffiliated with a national right-wing party. Similarly, in Columns (3) and (4) of Table 6we consider that the victories of mainstream left-wing and communist candidates, as well as of left-wing candidates unaffiliated with a national left-wing party, benefit the left-wing government.

As a robustness check, we report in an online Appendix (Table SA5, Appendix D) regressions which rely on the sample of the rulings where the CE did not use the vote differential to draw conclusions on the validity of the elections. Our regressions however only focus on the CE rulings which includes a discussion of the vote differential because only those decisions enable us to have a quasi-natural experiment.

It is worth pointing out that mayoral elections in the largest French towns are salient to national-level parties in France, and are even seen as tests of popularity for the current government. For instance, in the 2001 mayoral elections, the socialist (mainstream left-wing) party managed to win Paris and retain Lyons (respectively the first and second largest French cities), but lost many of the towns they controlled. In retrospect, it was said that these defeats heralded the 2002 presidential election where socialist candidate Lionel Jospin did not make it to the second round.

See Luchaire (1987) on the successive drafts of the Constitution of the Fifth Republic.

In an online Appendix (Table SA6, Appendix D), we include interaction variables between the vote differential and the affiliation of the election winners in our baseline specification. These regressions confirm the results in Columns (3) to (6) of Table 6 that the elections of mainstream-right wing, mainstream left-wing and communist politicians were likely to be upheld except when the vote differential was small. However, none of the interaction variables is significant, thus suggesting that the CE judges did not favor mainstream left-wing politicians at the expense of mainstream right-wing candidates, and vice versa.

The CE also acknowledged the existence of irregularities in the elections whose validity was questioned by the Préfets (in Columns 3 and 4 of Table 6 and in Columns 3 to 6 in Table SA5 in online Appendix D) and by the CNCCFP (in Columns 5 and 6 of Table SA5 in online Appendix D).

For the record, FN candidates won the mayoral elections in Orange, Marignane and Toulon in 1995 while FN candidate Cathérine Mégret also won the 1997 by-election in Vitrolles. She was reelected in 2001 but her victory was cancelled by the CE in its 28 January 2002 ruling n^○ 240196–240207. A reading of the ruling suggests that the CE judges were biased.

Between 1974 and 2007, Jean-Marie Le Pen was the FN candidate for all the presidential elections, except in 1981 because he did not receive the 500 signatures of elected officials, e.g., mayors, département and région council members, which are required to run in the presidential election. He received 0.75% of the votes in the first round of the 1974 election, 14.39% in 1988, 15% in 1995, 16.86% in 2002—when he made it to the second round of the elections—and 10.44% in 2007.

Between 1958 and 2007, the far-right was so much outside the French consensus that the mainstream right-wing parties systematically refused alliances with far-right candidates in local and in national elections (see Rémond (1982)’s seminal work on the different ideological groups in the French Right). Still, in the 2012 parliamentary elections, some mainstream right-wing candidates forged local alliances with far-right candidates. It is too early however to tell whether such alliances will normalize the situation of the far-right within French politics and will influence the judges' rulings in the long-run.

Most of the mayoral victories of Front National candidates occurred under right-wing governments, in June 1995, just after right-wing candidate Jacques Chirac’s victory against left-wing candidate Lionel Jospin in the May 1995 presidential elections, and later in a by-election in 1997. Consequently, the Winner Far Right variable has a positive (and not significant) coefficient in Columns 5 and 6 of Table 6and Table A1 when a right-wing government was in power after 1981, while it has a negative (and not significant) coefficient in Column 3 of Table A1, when a left-wing coalition governed France. These results can simply be viewed as statistical artifacts which reflect the fact that fewer mayoral elections were won by far-right candidates under left-wing than under right-wing governments.

Franck (2010) reports that the Conseil Constitutionnel judges demonstrated partisanship against Front National candidates when they ruled on the validity of parliamentary elections. But the Conseil Constitutionnel judges are political appointees, who are nominated by either mainstream left-wing or mainstream right-wing politicians and who serve a non-renewable term of 9 years. Their partisan behavior is therefore not surprising, unlike that of the CE judges who are high-level civil servants.

During the 1958–2007 period, many CE judges were known sympathizers of some political parties. Some even left the CE to pursue a political career. As a result, the most well-known CE member from the left is Prime Minister Laurent Fabius (1984–1986). Among the CE members who hailed from the right, the most famous are Prime Minister Michel Debré (1959–1962), Prime Minister Edouard Balladur (1993–1995) and Prime Minister Dominique de Villepin (2005–2007). Conversely, as far as we could tell, only three CE judges were known FN sympathizers—Alain de Lacoste-Lareymondie, Michel Massenet and Pascal Arrighi de Casanova. Of course, none of these three CE members had a significant political career. It is also worth noting that few CE judges were known members of the communist party, with the notable exceptions of Jean Kahn and Guy Braibant. However, this need not have hurt the elections of communist politicians since the communists were junior partners in the left-wing coalition headed by the Socialist Party between 1981 and 1984, and again between 1997 and 2002.