Thread by @DawieScholtz

1/ So my thoughts on Gauteng: TOO CLOSE TO CALL

By far the most interesting and important one to watch during this election.

With my best-guess assumptions:
ANC - 48
DA - 31
EFF - 12

ANC biggest for sure, but it's close between ANC and 50. Could go either way.

2/ Let's start by looking at recent electoral history in Gauteng (ANC-DA-EFF):

2011: 60-33-0
2014: 55-29-10
2016: 46(!)-37-11

So the ANC went-14% between 2011 and 2016. Two reasons: Turnout and massive township losses.

3/ Turnout is one of the reasons why the ANC is losing % in GP. The DA base turns out like crazy, the ANC base doesn't and I think it'll happen again this year. Turnout % (suburban-township):

2011: 66-54
2014: 81-74
2016: 71-53

That 2016 gap of 18% was HUGE. DA needs 10%+

4/ Furthermore, the ANC's losing tons of support in the townships since 2011. the Gauteng black electorate (ANC-DA-EFF):

2011: 86-7-0
2014: 75-7-13
2016: 69-10-16

So that's -17% for the ANC in 5 years. Extraordinary finding. And terrifying for the ANC!

5/ The big question in this election is whether Cyril Ramaphosa can pull back some of these losses. By-elections and polling suggests there's an outside possibility he can, but probably not. We shall see.

This point of uncertainty is the key question in this election.

6/ Something should also be said about the split between Northern and Southern Gauteng townships. The ANC is stacking up big losses up north in by-elections the last year, but is more resilient down South.

7/ All-in-all, Gauteng is going to be insanely close I think!

The outcome will ultimately depend on the relative turnout of each party's base and Cyril Ramaphosa's ability to claw back support (or not?) that was lost in the townships 2011-2016.

It's going to be fascinating!

8/ DISCLAIMER: This is based on unscientific 'best-guesses' for various unknown variables, informed by polling and by-elections. Not a formal prediction.