TABLE 2
FROM:
Forecasting the 2007 French Presidential Election: Ségolène Royal and the Iowa Model
Michael S Lewis-Beck, Éric Bélanger and Christine Fauvelle-Aymar
BACK TO ARTICLETable 2. 2007 French presidential election forecasts, from different forecastersa
| Forecaster | Forecast, first round | Forecast, second round |
|---|---|---|
| AA | 49.4% for left (3)b | |
| BFL | 46.4% for left (3) | 51.0% for left (5) |
| JJS | 44% for left (2) | 47.1% for left (1)b |
| LFA | 48.6% for left (2)b | |
| LL | 39.6% for left (1) | <50% for left (4)b , c |
| JE | 16.7 % for Le Pen |
| AA | 15.6% for Le Pen |
| JJS | 16.0% for Le Pen |
a The forecasters are listed alphabetically, and indicated by letters: AA=Auberger (2007), BFL=the regional model of Bélanger et al. (2007), JJS=Jérôme and Jérôme-Spéziari (2007), LFA=Lafay et al. (2007), LL=Lemennicier and Lescieux (2007).
b =correctly predicted the presidential winner.
( )=the figures in parentheses in the table indicate the relative rank of the forecast, comparing its success to the others. For example, (1) indicates the most accurate forecast.
c LL did not make a point forecast for the second round, rather only a win-lose forecast, calling a loss for the left. Further, the second-round forecast is based on visual inspection of the pre-election distribution of vote intention in a survey 1 month before the first round, while the first-round forecast is based on a multiple regression equation.
