Continuing our deep dive on the outcome of our August poll and it’s implications for the December elections, we now turn our attention to the 12 regions where our model downgraded Mahama’s chances. 

The unlucky dozen: 

  • Ahafo – Mahama saw a slight dip in his chances here from 24.4% to 21%. Our model continues to view this region as a bit of a long shot for Mahama and a fairly easy win for Nana Addo. 

  • Bono – Much like the Ahafo region, Mahama’s chances here have dropped from 26.4% to 24%. Nana Addo also saw a dip in his chances here as the model thinks the chances of a tie between the too should be higher than previously estimated. Still Nana Addo has a slight edge at 55% chance of winning.

  • Bono East – Definitely the biggest surprise after updating the forecast with the August poll. Our model counts this region as an NDC stronghold. That’s why it was a shocker for us to see Mahama’s chances drop from 87.2% to 74%. He’s still favorites to win the region but that steep of a drop suggests the race is far from settled.

  • Central – The race in this key swing region continues to tighten. Mahama’s chances dropped from 43.4% to 37% but the chance of a tie doubled from 8% to 16%. The race here is effectively a toss up. 

  • Eastern – Mahama’s chances dropping in an NPP stronghold may not surprise you. However, considering Nana Addo saw a bigger drop in his chances here, it stands to reason that the NPP still has some work to do convincing its own base. Still Mahama’s 4% chance here is a really long shot bid and barely has any impact on his overall chances at the National level. 

  • Greater Accra – Like Central Region, the forecast in the capital region is effectively a toss up. Both Mahama and Nana Addo saw equal drops in their chance here while the probability of a tie now sits at 18%, up from 6% prior to the update. As of this writing, our model gives Mahama a 40% chance of winning, while Nana Addo is at a 41% chance.

  • Northern – Another big surprise coming out of an NDC stronghold. Mahama’s chances dropped from 86.8% to 75%. The model still expects him to win the region but with a 20% chance of a tie it also expects the NPP to make a difference if it were able to mount a powerful challenge to the NDC’s dominance.

  • Oti – Like Northern and Bono East, Mahama’s chances dropped yet again in an NDC stronghold. At 82% our model expects him to still win Oti. Prior to the update, our model gave him almost 97% chance of winning the region.

  • Savannah – Starting to see a trend here with the dips in Mahama’s chances in NDC strongholds. Prior to the update, our model forecasted his chances at 94.5%. Today, it has dropped to 80%. 

  • Upper West – Much like Savannah, Mahama saw his chances in this NDC stronghold drop from 92% to 82%. He’s still expected to win the region but the model thinks the chance of a tie should not be discounted in some of the more extreme scenarios.

  • Western – Before the forecast, our model gave Mahama long odds of winning this swing region. His forecasted chances stood at 12.8%. After our August poll, it’s now at 11%. Indeed the Western Region is one of only 2 in our forecast that became slightly less competitive after the August poll. Our model favors Nana Addo to win the region.

  • Western North – Rounding out the trend in Mahama’s falling chances in NDC strongholds, the Western North forecast saw a 14.5% drop from 96.5% to 82%. Our model expects him to win Western North with roughly the same odds as winning Oti and Savannah. 

One nuance to bear in mind with the forecast  is that a decrease in a candidate’s chances of winning is not entirely bad if the shift moves towards making the race competitive in that candidate’s favor. We see this in the case of Mahama. Almost all of the drops in his chances of winning came from gains in the chances of a tie, which benefit him as the challenger in the elections. Normally our model considers an increase in the chance of a tie first before increasing the chances of the opposing candidate. So if Mahama’s chances fall in a region that is likely to win, then the model gives the bounce to the chance of a tie before considering an increase in Nana Addo’s chances.

%d bloggers like this: