I was a little late to Kos’ thread (which mostly I strongly agree with) and I think it’s worth looking at the details of favorability ratings a little more closely. For the purposes of using apples vs apples (to the degree that I can), i’m going to refer to the NY Times data included here (covering 1976 - 2012, fivethirtyeight.blogs.nytimes.com/...). Unfortunately, I do not have recent data from that specific poll, so i’m using the collection of data and using the average for favorability, but the high numbers for undecideds (All from HuffPost). This gives us the best and worst case scenario for each candidate.
The Favorables
First, let’s deal with Kos’ specific point.
“Check out the chart below: see that inflection point when Clinton’s unfavorables overtook her favorables? That was April 2015. You know what month Bernie Sanders entered the campaign? April 2015. Most will come home after Sanders bows out.”
OK, I don’t agree with Kos’ rather large assumption that most will come home once Sanders comes home. Some of these are going to be Rs and Is that just started paying attention again and remembered why they disliked her. But that issue aside, in April, 2015, Hillary’s un-favorability was around 45% with 15% or less undecided. Even if we assume she can work her way back to that rating, here are some facts:
- The only Presidential candidate in the last 40 years with a higher un-favorability rating than 45% was Carter in 1980. He lost big.
- The only two candidates in the last 40 years with less than 15% undecided, according to the NY Times (which admittedly allows for higher undecideds), is Carter in 1980 and Ford in 1976. They both lost.
No non-incumbent Presidential nominee in the last 40 years has had a favorability greater than 37%. The highest un-favorability (at this point in the election) for a candidate that won an election was Al Gore in 2000 (Scalia not-withstanding) who had a 36% unfavorable rating. The next closest is Bush Sr with 35%. Hillary’s unfavorable rating is 9-10 points higher than ANYONE that has run for President as a non-incumbent in the last 40 years and won.
For incumbents, Obama had the highest un-favorability rating of anyone at this point to go on and win a Presidential election within the last 40 years when he had a 42% negative in 2012. That remains 3-12% points lower than Hillary regardless of whether you choose to compare the more important current data or cherry pick data from 9 months ago.
Now, of course Hillary’s approval rating today, which is more directly comparable to the data in the 538 charts cited above, is a 54% disapproval rating with 6% or less undecided (in all but one poll). Again, Carter had a 58% unfavorable in 1980 and leads in that category. In any case, no Presidential candidate in history has had that few undecideds. Even recent incumbents Clinton, Bush, and Obama had 16-21% undecideds.
The Undecided
Undecideds are important. It’s more difficult to move people who are already negative about you than those that are undecided or positive. There are 3 elections in the last 40 years where at this point in the election, the winning candidate had a higher negative favorability rating than their opponents:
-
1988: Bush (35% unfavorable, 31%undecided) defeated Dukakis (16% unfavorable, 50% undecided)
-
2004: Bush Jr (41% unfavorable, 19% undecided) defeated Kerry (27% unfavorable, 45% undecided)
-
2012: Obama (42% unfavorable, 17% undecided) defeated Romney (37% unfavorable, 37% undecided)
If you are interested in that disputed election of 2000, here are the numbers:
-
2000: Gore (36% unfavorable, 32% undecided) defeated Bush (31% unfavorable, 32% undecided)
Nobody has ever defeated an opponent with lower negatives at this point in an election in the last 40 years - unless there are a huge number of undecided (the closest was Gore which we all now was silly close). Now compare that to our two Democratic candidates this year:
-
Clinton: 54% unfavorable, 6% undecided
-
Sanders: 38% unfavorable, 8-23% undecided
Compare both with Clinton’s data as of last April (cherry pick the data):
-
Clinton (April, 2015): 45% unfavorable, <25% undecided (best guess on undecideds here)
Obviously, Bernie has more undecideds than Hillary but it’s nowhere close to what he would need to have to end up at 54% negative. Now let’s look at the Republican side (since high undecideds on either side can swing an election).
-
Trump: 58% unfavorable, 3-13% undecided
-
Cruz: 47% unfavorable, 13-29% undecided
-
Rubio: 40% unfavorable, 17-37% undecided
-
Kasich: 33% unfavorable, 33-60% undecided
This basically plays out how you would expect it would. As Kos noted, Hillary would likely beat Trump. Neither of them are going to shift a lot in this election. Hillary and Cruz are close enough though that I would not guarantee it. Cruz has enough undecideds to swing it with Hillary or vice versa. Bernie, on the other hand, is going to crush Trump. There simply aren’t enough undecideds to move that election the other way. And he also likely cruises (Cruzes?) past Cruz. The high undecideds give both of our candidates a better chance with Rubio and Kasich, but Bernie is in a much better position to survive that.
Conclusion
The good news is that I expect either Hillary or Bernie can and will defeat Trump. The bad news is that if we want coattails, we need Bernie and he is a bit of a long-shot to win the primary — as of right now anyhow. The worse news is that Rubio could spell trouble if he can get back into this.
This is not exhaustive. There are other ways to look at this. But I think from everything I’ve seen, this is pretty consistent with all of the other data points such as head-to-head data and it takes into account that predictability of favorability ratings throughout history.
Just keep those undecideds in mind in the head-to-head match-ups. It matters that Bernie is winning against Trump, but Bernie vs. Rubio can move significantly. That’s where the head-to-head data could be off. Everywhere else it’s pretty much dead on and predictive.
Sources
fivethirtyeight.blogs.nytimes.com/…
elections.huffingtonpost.com/...