Skip to main content
Log in

Forecasting the 2012 French presidential election: Comparing vote function simulations and vote intention polls

  • Symposium Article
  • Published:
French Politics Aims and scope

Abstract

In this article, we build a pooled time-series vote function for the 2012 French presidential election both at regional and national levels, taking into account our 2002 and 2007 past experiences. Particularly, a variable capturing the behaviour of François Bayrou’s voters has been included. From 2009 (Q1) up to 2011 (Q4), we have used the vote function in order to simulate the votes for the Left, the Right and the National Front, at the two rounds, asking ‘what would be the result of the vote if elections were held now’? Then we compare the vote simulations to the vote intention polls and we try to explain the differences between the two predicting methods. The Left and François Hollande are largely and continuously declared as the winners by the pollsters since 2009. On the contrary, the vote function simulations result in a final forecast giving Nicolas Sarkozy 50.6 per cent at the second round. Maintaining this slight advantage (while fixing Incumbent's popularity at a 32 per cent approval rate) requires the unemployment rate (U) not to exceed 9.6 per cent 4 months before the election date.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Figure 1
Figure 2
Figure 3
Figure 4

Similar content being viewed by others

Notes

  1. For a general review, see Lewis-Beck (2005).

  2. See Lafay et al (2007), Lemennicier-Bucquet et al (2010), Lewis-Beck et al (2008).

  3. This event has is related to S. Royal’ s own strategy asking F. Bayrou to rally her at the second round, whereas the first round was not yet played. These two parameters incited opportunistic (and/or strategic) moderate socialists to support Bayrou instead of Royal at the first round. This could explain why the F. Bayrou vote grew up from 12 per cent to 18 per cent and correlatively why total Left score was so low.

  4. See Jérôme et al (2003).

  5. See Jerome and Jerome-Speziari (2004).

  6. Jérôme et al (1993) test a purely economy model and then a purely political economy model for the legislative elections.

  7. Geographical models are often criticized for been unable to predict the brutal structural changes in the voting structure. Further, these models are raising true methodological problems when transferring at the individual level the correlations achieved at the territorial level. Comparison and control by way of surveys at the individual level are often required.

  8. Baltagli (1995).

  9. This method has the disadvantage, however, of combining inference and estimation problems, because it is exposed to the risk of heteroscedasticity of the cross sectional data and to the problems of serial correlation of the residual of the time-series data. These difficulties– where necessary -can be rectified by using tests such as a covariance model (for example, a dummy variable least squares model), where each region and each period are coded next to the explanatory variables. In our model, dummies targeting local situations or specific strength areas play a similar role.

  10. However, in 2002, L. Jospin has been eliminated on the first round such as G. Deferre or J. Duclos in 1969. Then, in order to obtain a continuous variable for the Right, we forecasted the score J. Chirac would have had if L. Jospin had been qualified for the second round.

  11. See Jérôme and Jérôme-Speziari (2011).

  12. In France, long run empirical studies show that economy, compared with domestic politics and international concerns, influences, at least, 50 per cent of the Popularity index.

  13. Data for the Right in 2002 are regional simulations coming from an experiment (Jérôme and Jérôme-Speziari, 2004) initially computed for the incumbent Left at the second round. It yields a vote combination of 46 per cent at a national level. The Left would have obtained an absolute majority only in Midi-Pyrénées. Of course, this means that J. Chirac would have won with a 54 per cent share if is opponent would have been L. Jospin instead of J-M. Le Pen.

References

  • Alesina, A. (1987) Macroeconomic policy in a two-party system as a repeated games. Quarterly Journal of Economics 102 (3): 651–678.

    Article  Google Scholar 

  • Baltagli, B.H. (1995) Econometric Analysis of Panel Data. New York: John Wiley and Sons.

    Google Scholar 

  • Blais, A., Nadeau, R., Gidengil, E. and Nevitte, N. (2001) Measuring strategic voting in multiparty plurality elections. Electoral Studies 20 (3): 343–352.

    Article  Google Scholar 

  • Campbell, J.E. (1992) Forecasting the presidential vote in the states. American Journal of Political Science 36 (2): 386–407.

    Article  Google Scholar 

  • Goguel, F. (1981) Chroniques électorales, 3 tomes. Paris, France: Presses de la FNSP.

    Google Scholar 

  • Hibbs, D.A. (1987) The Political Economy of Industrial Democracies. Cambridge, US: Harvard University Press.

    Book  Google Scholar 

  • Holbrook, T.M. (1991) Presidential elections in space and time. American Journal of Political Science 35 (1): 91–109.

    Article  Google Scholar 

  • Jérôme, B. and Jérôme-Speziari, V. (2001) La France réélirait Chirac. in Ils ont trouvé la formule qui donne le résultat des élections, L’Expansion, n656, 8 au 21 novembre 2001, pp.74–75.

  • Jérôme, B. and Jérôme-Speziari, V. (2003) A Le Pen vote function for the 2002 presidential election: A way to reduce uncertainty. French Politics 1 (5): 247–251.

    Article  Google Scholar 

  • Jérôme, B. and Jérôme-Speziari, V. (2004) La modélisation des effets du changement de calendrier électoral: une approche politico-économique. Dans Les systèmes électoraux: permanence et innovation, dir. A. Laurent et H.-P. Frognier. Paris: l’Harmattan.

  • Jérôme, B. and Jérôme-Speziari, V. (2007) Prévision de la présidentielle 2007, qui sera aux commandes du pays? in Lechypre E. ‘ France, l’Equation à 7 inconnues’, L’Expansion, janvier, n°715, p.28.

  • Jérôme, B. and Jérôme-Speziari, V. (2010) Fonctions de vote et prévisions électorales, une application à la présidentielle de 2007. Canadian Journal of Political Science 43 (1): 163–186.

    Article  Google Scholar 

  • Jérôme, B. and Jérôme-Speziari, V. (2011) Nowcasting Versus Polls, An Application to the French Presidential Election. IRGEI-LARGEPA, Paris 2 Pantheon-Assas. Working Paper.

  • Jérôme, B., Jérôme-Speziari, V. and Lewis-Beck, M.S. (2001) Évaluation économique et vote en France et en Allemagne. In: D. Reynié and B. Cautres (eds.) L’Opinion européenne 2001. Paris, France: Presses de sciences Po, pp. 101–122.

    Google Scholar 

  • Jérôme, B., Jérôme-Speziari, V. and Lewis-Beck, M.S. (2003) Reordering the French election calendar: Forecasting the consequences for 2002. European Journal of Political Research 42 (3): 425–440.

    Article  Google Scholar 

  • Jérôme, B., Lewis-Beck, M.S. and Lafay, J.-D. (1993) In: J.-D. Lafay, Les prévisions des modèles politico-économiques. Le Figaro, 19 mars.

  • Key, V.O. (1966) The Responsible Electorate. New York: Vintage.

    Book  Google Scholar 

  • Lafay, J.-D., Facchini, F. and Auberger, A. (2007) Modèles politico-économétriques et prévisions électorales pour mai 2007. Revue Française d’Economie XXI (4): 145–164.

    Article  Google Scholar 

  • Lemennicier-Bucquet, B., Lescieux-Katir, H. and Grofman, B. (2010) The 2007 French presidential election. Canadian Journal of Political Science 43 (1): 137–161.

    Article  Google Scholar 

  • Lewis-Beck, M.S. (1985) Un modèle de prévision des élections législatives françaises avec une application pour 1986. Revue française de science politique 35 (6): 1080–1091.

    Article  Google Scholar 

  • Lewis-Beck, M.S. (2005) Election forecasting: Principles and practice. British Journal of Politics and International Relations 7 (2): 145–164.

    Article  Google Scholar 

  • Lewis-Beck, M., Bélanger, E. and Fauvelle-Aymar, C. (2008) Forecasting the 2007 French presidential election: Ségolène Royal and the Iowa model. French Politics 6 (2): 106–115.

    Article  Google Scholar 

  • Lewis-Beck, M.S. and Jérôme, B. (2010) Special issue, European economic forecasting. International Journal of Forecasting, n°1, January–March (26) Choice 59 (1): 61–82.

    Google Scholar 

  • Lewis-Beck, M., Nadeau, R. and Bélanger, E. (2010) Forecasting the vote for the third party: The British Liberals, 1974–2004. British Journal Of Politics and International Relations 12 (4): 634–643.

    Article  Google Scholar 

  • Lewis-Beck, M., Nadeau, R. and Bélanger, E. (2011) Now casting V. polling: The 2010 UK election trials. Electoral Studies 30 (2): 284–287.

    Article  Google Scholar 

  • Mingat, A. and Salmon, P. (1988) Alterable electorates in the context of residential mobility. Public Choice 59 (1): 67–82.

    Article  Google Scholar 

  • Mueller, D.C. 2003 Public Choice III. Cambridge, UK: Cambridge University Press.

    Book  Google Scholar 

  • Nannestad, P. and Paldam, M. (1994) The VP-function: A survey of the literature on vote and popularity functions after 25 years. Public Choice 79 (3–4): 213–245.

    Article  Google Scholar 

  • Rosa, J.-J. and Amson, D. (1976) Conditions économiques et élections: une analyse Politicoéconomique 1920–1973. Revue française de science politique 26 (6): 1101–1124.

    Article  Google Scholar 

  • Rosenstone, S.J. (1983) Forecasting Presidential Elections. New Haven et London: Yale University Press.

    Google Scholar 

  • Siegfried, A. (1913) Tableau Politique de la France de l’Ouest sous la troisième République. Paris, France: Armand Colin.

    Google Scholar 

  • Tiebout, C.M. (1956) A pure theory of local expenditures. Journal of Political Economy 64 (octobre): 416–424.

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Appendices

Appendix A

Specification of variables

Vote function for the incumbent (equation 1)

Explained variable

VPT1 :

The combination of votes in regions where candidates are close to the outgoing majority. The Right is the outgoing majority in 1974, 1981, 1988, 1995 and in 2007. The Left is the outgoing majority in 2002.

Explanatory variables

ΔU:

Variation of the regional rate of unemployment between December in year (t) preceding the presidential election and December in year (t−1).

VLG :

Votes for the candidates who are close to the outgoing majority at the past legislative elections (1973, 1978, 1986, 1993, 1997 and 2002).

ZFA :

Variable scored 1 or (−1) (zero otherwise) in each regional strength area at the presidential election. (−1) is for the stronghold of the opposition and 1 is for the stronghold of the majority. We call absolute strength areas, regions that – at the period of the election 1981, 1988, 1995, 2002 and 2007 – have never changed political colour since 1974. (See appendix A for the details).

ZFDRCENT:

Variable scored 1 (zero otherwise) for the Center (now MODEM – Mouvement des Démocrates) in each regional strength area shared with the Right at the presidential election.

Alsace, Bourgogne, Centre, Lorraine, Basse-Normandie, Île-de-France, Pays de la Loire, Rhône-Alpes.

ZFGACENT:

Variable scored 1 (zero otherwise) for the Center (now MODEM – Mouvement Des DÉMocrates) in each regional strength area shared with the Left at the presidential election. Aquitaine, Auvergne, Bretagne, Limousin, Midi-Pyrénées, Nord-Pas-de-Calais (NPDC), Poitou-Charentes.

COPMPT :

Gap between the quarterly popularity (IFOP) of the Prime Minister and the President of the Republic measured 6 months before the election in case of COHAB (1988, 1995, 2002).

POPT :

Quarterly popularity (IFOP) of the President (or the Prime minister in case of COHAB) measured 6 months before the election, when it takes place at the expected date. The election in 1974 was an unexpected one, caused by the death of President G. Pompidou, so this variable scored zero at this date.

INCLEFT :

Variable equal to 1 when the Left is the outgoing majority – that is, only in 2002 – (zero otherwise).

CORS07 :

Dummy variables scored 1 for Corsica in 2007 (zero otherwise), due to the off-norm scores of the Right in this region

Vote function for the Right at the second round (equation 2)

Explained variable

VPTD2 :

Combination of regional votes for the candidate of the Right (per cent). The Right is the outgoing majority in 1974, 1981, 1988, 1995, 2007 and is in the opposition in 2002.Footnote 13

Explanatory variables

VPTD1 :

Combination of regional votes for the candidates of the Right block (per cent).

ZFND :

Dummy variable scored 1 (zero otherwise) for the Right in strength areas shared with the FN at the second round. This is the case for Alsace, Centre, Champagne-Ardennes, Franche-Comté, Lorraine, PACA, Rhône-Alpes.

ZFNG :

Dummy variable scored 1 (zero otherwise) for the Left in strength areas shared with the FN at the second round. This is the case for Haute-Normandie, NPDC, Picardie.

COHAB :

Dummy scored 1 (zero otherwise) in1988 and in 1995. This variable calculates the decline for the candidate of the Right at the second round, when the Right is the ruling coalition with a President belonging to the Left.

ZCENTD:

Variable scored 1 (zero otherwise) for the center or MODEM in each regional strength area shared with the Right at the presidential election. This variable calculates the electoral loss for the Right in rightist regions when leftist voters (mostly socialists) who have supported MODEM and F. Bayrou at the first round go back to their initial camp.

ZCENTG:

Variable scored 1 (zero otherwise) for the Center or MODEM in each regional strength area shared with the Left at the presidential election. This variable calculates the electoral loss for the Right in leftist regions when leftist voters (mostly socialists) who have supported MODEM and F. Bayrou at the first round go back to their initial camp.

Vote Function for the FN (equation 3)

Explained variable

VFNP1 :

Votes (per cent) obtained by the FN at the first round of the presidential election.

Explanatory variables

VFNLG1 :

Votes (per cent) obtained by the FN at the first round of the previous legislative elections (1986, 1993, 1997, 2002).

ΔU:

Variation of the regional rate of unemployment between December in year (t) preceding the presidential election and December in year (t−1).

ZSARKFN07 :

Dummy scored 1 in 2007 (0 otherwise) in regions where N. Sarkozy has operated an electoral puncture upon FN vote. (Alsace, Bourgogne, Centre, Franche-Comté, Languedoc-Roussillon, Haute-Normandie, PACA, Rhône-Alpes).

FNWEAK :

Dummy scored 1 (zero otherwise) in regions where the result for the FN is traditionally below its national result on the average. (Aquitaine, Auvergne, Bretagne, Limousin, Midi-Pyrénées, Basse-Normandie, Île-de-France, Pays de la Loire and Poitou-Charentes).

CORS02 :

Dummy scored 1 (zero otherwise) for Corse in 2002.

Appendix B

Opposition and majority strongholds

In 1974

Absolute strength areas for the Right (the majority)

Alsace, Aquitaine, Auvergne, Bretagne, Centre, Champagne-Ardennes,

Franche-Comté, Lorraine, Basse-Normandie, Île-de-France, Pays de la Loire,

Poitou-Charentes, Rhône-Alpes, Corse.

Absolute strength areas for the Left (the opposition)

Bourgogne, Languedoc-Roussillon, Limousin, Midi-Pyrénées, Nord-

Pas-de-Calais (NPDC), Haute-Normandie, Picardie and Provence-Alpes-

Côte-d’Azur (PACA).

In 1981

Absolute strength areas for the Right (the majority)

Alsace, Auvergne, Bretagne, Centre, Lorraine, Basse-Normandie, Pays de la Loire, Corse.

Absolute strength areas for the Left (the opposition)

Bourgogne, Languedoc-Roussillon, Limousin, Midi-Pyrénées, NPDC,

Haute-Normandie, Picardie, PACA.

In 1988

Absolute strength areas for the Right (the majority)

Alsace, Centre, Lorraine, Basse-Normandie, Pays de la Loire, Corse.

Absolute strength areas for the Left (the opposition)

Bourgogne, Languedoc-Roussillon, Limousin, Midi-Pyrénées, NPDC,

Haute-Normandie, Picardie.

In 1995

Absolute strength areas for the Right (the majority)

Alsace, Centre, Lorraine, Basse-Normandie, Pays de la Loire, Corse.

Absolute strength areas for the Left (the opposition)

None

In 2002 and 2007

Absolute strength areas for the Right (the opposition)

Alsace, Centre, Lorraine, Basse-Normandie, Pays de la Loire, Corse.

Absolute strength areas for the Left (the majority)

None

Rights and permissions

Reprints and permissions

About this article

Cite this article

Jerôme, B., Jerôme-Speziari, V. Forecasting the 2012 French presidential election: Comparing vote function simulations and vote intention polls. Fr Polit 10, 22–43 (2012). https://doi.org/10.1057/fp.2011.22

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1057/fp.2011.22

Keywords

Navigation