February 24, 2016 10:15 AM
Hillary Clinton is poised for a pretty big win in Georgia and in several other primaries next week. But how much will she win by? How much does she need to win by?
If you’re interested in following the polls, you’re probably already reading Nate Silver’s FiveThirtyEight. Before Obama was reelected four years ago, many political commentators were dismissive of Silver’s analyses, but maybe more people understand the general methodology by now.
According to FiveThirtyEight’s “polls-only” forecast for Georgia’s Democratic primary, Clinton has a 99 percent chance of victory. As I write this, the numbers most heavily weighted in the FiveThirtyEight analysis are from a poll by Landmark/RosettaStone that has Clinton beating Bernie Sanders by a staggering 72.0 to 19.8 percent, with 8.1 percent undecided. In that poll, Clinton has a +40 edge even among white voters, a demographic that she apparently lost in Nevada. She has a +55 edge among black voters in that Landmark poll. (By the way, I’m personally sympathetic to the arguments that Sanders’s positions align much better than Clinton’s with the positions of African-American voters, but the Sanders message simply has not taken hold in the way his supporters expected.)
But let’s not get caught up in just the latest poll. The key to FiveThirtyEight’s methodology is to look at all the credible polling. Based on the data available so far, Clinton is projected to get 68.7 percent of next week’s primary vote, with Sanders far behind at 27.5 percent.
FiveThirtyEight has this year also rolled out “polls-plus” forecasts, which adjust the range of outcomes based on state polling, national polling, and endorsements. Using that data, Clinton would likely win by even more — 73.1 to 23.2. Those numbers will move around somewhat over the next 5 days, but I doubt by much.
Public Policy Polling recently showed Clinton overwhelmingly ahead (+23 or more) in other early March primaries: Alabama, Arkansas, Louisiana, Mississippi, Tennessee, Texas, and Virginia.
Given the current expectations, it would be a moral victory for Sanders if he even reached 40 percent of the Democratic primary vote in Georgia. But would it matter if Sanders took 40 percent instead of 30 percent? Probably not. With big wins in so many states, Clinton could build an insurmountable lead in delegates. Sanders would need some very unlikely and truly huge wins in later primaries to make up the necessary ground.
According to Frontloading HQ, Georgia will send 116 delegates to the convention — 22 at-large delegates, 13 pledged elected officials, 67 district delegates, and 14 unpledged “super delegates.”
If you start doing the math, huge differences reveal themselves if Clinton gets 60 percent or if she gets 70 percent on Tuesday, but, for the most part, the number of pledged delegates will reflect the proportions of the statewide vote.
Let’s say Clinton takes 60 percent: she would get 13 at-large delegates and 8 pledged elected officials for her statewide margin, but she might only get about 38 of the district delegates; several districts assign four delegates, and a 60-40 margin would translate into a 2-2 split. So that could give Clinton 59 and Sanders 43 of the 102 pledged delegates — not counting the super delegates.
But if Clinton hits 70 percent, those numbers change significantly. She’d take 15 at-large delegates, 9 pledged elected officials, and somewhere around 47 of the district delegates (districts with 4 delegates would be split 3-1). That could give Clinton around 71 delegates to 31 for Sanders, before the 14 super delegates are included.
So will Clinton lead Sanders by as few as 16 pledged delegates in Georgia? Or will her margin be somewhere around 40 — or even more?
Even if the math is against him, Sanders could make a strong case for remaining in the race if he hits that 40 percent mark across the South, and there would be good reasons for him to stay in the race no matter what happens.
Now, could all this conventional wisdom — and all the extensive polling data — be totally wrong? Could something happen in the next few days that would upend the likely outcome in South Carolina on Saturday and lead to a Sanders surge next Tuesday? I know plenty of Sanders supporters who seem to be counting on that happening, somehow, but the likelihood of such a dramatic shift is quickly fading.