When the polls closed, while the pundits pundited, I was making an election night wish list on Google Sheets. What would be my dream result? I had races on there from President to local ballot initiative, but the sheet would soon turn depressingly red. So many targets — some that were always unrealistic, some that had seemed perfectly realistic — left unmet.
I had included all battleground US Senate races, of course, and those that had fallen off the battleground list during the campaign as well. Same for the gubernatorial races. Of the lot of them — 22 in total — I got to format just 8 in blue by the end. And two of those involved having Jim Justice winning in West-Virginia while also leaving at least 4% for the Mountain Party.
But how are we to evaluate all those losses, and some of those wins? What lessons can we draw from them? Was it because of candidate quality (ours or the opposing side's), campaign quality, demographics? I'd like to hear your answers, but first we need benchmarks. And I can think of three.
First: how well did each Democrat who ran for the Senate or for Governor perform in comparison to Hillary Clinton? Did they run ahead of her? Did they do worse than her, even as she lost her own race? I will include every candidate whose race looked at any time like it could be competitive:
Outperforming - or underperforming - Clinton
State |
DEM CANDIDATE |
GOP CANDIDATE |
DEM MARGIN |
Clinton margin |
Difference |
|
US SENATE |
|
|
|
|
ILLINOIS |
Duckworth |
Kirk |
14.2 |
16.6 |
-2.4 |
Colorado |
Bennett |
Glenn |
3.9 |
4.2 |
-0.3 |
Nevada |
Masto |
Heck |
2.4 |
2.4 |
0.0 |
New Hampshire |
Hassan |
Ayotte |
0.1 |
0.4 |
-0.3 |
Pennsylvania |
McGinty |
Toomey |
-1.7 |
-1.1 |
-0.6 |
Missouri |
Kander |
Blunt |
-3.2 |
-19.0 |
15.8 |
Wisconsin |
Feingold |
Johnson |
-3.4 |
-0.9 |
-2.5 |
North Carolina |
Ross |
Burr |
-5.8 |
-3.8 |
-2.0 |
Florida |
Murphy |
Rubio |
-7.7 |
-1.2 |
-6.5 |
Indiana |
Bayh |
Young |
-9.7 |
-19.0 |
9.3 |
Arizona |
Kirkpatrick |
McCain |
-12.3 |
-3.9 |
-8.4 |
Georgia |
Barksdale |
Isakson |
-14.2 |
-5.3 |
-8.9 |
Ohio |
Strickland |
Portman |
-21.4 |
-8.5 |
-12.9 |
iowa |
Judge |
Grassley |
-24.5 |
-9.5 |
-15.0 |
|
GOVERNOR |
|
|
|
|
West-Virginia |
Justice |
Cole |
6.8 |
-42.2 |
49.0 |
Montana |
Bullock |
Gianforte |
3.8 |
-20.5 |
24.3 |
North Carolina |
Cooper |
McCrory |
0.1 |
-3.8 |
3.9 |
New Hampshire |
Van Ostern |
Sununu |
-2.3 |
0.4 |
-2.7 |
Missouri |
Koster |
Greitens |
-5.9 |
-19.0 |
13.1 |
Indiana |
Gregg |
Holcomb |
-6.0 |
-19.0 |
13.0 |
Vermont |
Minter |
Scott |
-8.7 |
26.5 |
-35.2 |
Data sources: NY Times (Senate), Politico (Governor), 2016 National Popular Vote Tracker by David Wasserman (President)
Possibly surprising fact: 11 of the 14 Democratic nominees for Senate who were, at least at some point, touted as credible nominees did worse than Hillary Clinton. At least if you use this measure (the winning margin between Republican and Democrat). Catherine Cortez Masto, in Nevada, did exactly as well as Clinton. Just two of these Democrats did better; but they did a lot better: Jason Kander and Evan Bayh.
It works out a little bit differently if you go by Democratic share of the vote (because of the relatively high percentage going to Johnson and Stein in the presidential race), but only a little. If you look at it that way, Maggie Hassan and Michael Bennett pull ahead of Clinton, but Masto falls behind her.
So much for Hillary being a flawed candidate who dragged down Democratic prospects? Maybe, but you can also turn the comparison around. 11 of the 14 Republican candidates in these races got a better margin than Donald Trump did. As ubiquitous as the analyses this week are about how Trump turned out to have unique strengths that explain why he won, he still ran behind his downballot Republican colleagues in the Senate races in most of these states. Which is why the National Review's Dan McLaughlin crowed that “This was a big night for Republicans and a terrible one for Democrats, and that extends far beyond the personalities of Donald Trump and Hillary Clinton.”
The picture looks very different in the gubernatorial races. Here, most Democrats ran well ahead of Clinton. In West-Virginia, mining magnate Jim Justice pulled off a comfortable 7-point win even as Clinton lost to Trump by an unprecedented, lopsided margin of 27% to 69%. The 50-point difference between their margins is easily the biggest you'll find in this diary. But it wasn't just him. Steve Bullock, incumbent Governor of Montana, got reelected by a margin that was 24 points better than Clinton's. Chris Koster in Missouri and John Gregg in Indiana both lost, but both by 13 points less than Clinton.
So what did Kander, Bayh, Justice, Koster and Gregg have going for them that Clinton didn't? Do they have lessons to offer for national Democrats? There are always specificities, of course. Bayh had his family name, and a jealously-guarded campaign funding war chest from yesteryear. Justice, Koster and Gregg benefited from running for Governor rather than Senator, when ticket-splitting has become much more rare in federal races. Nevertheless it might be useful to look at these five politicians — and especially Jason Kander — as the Democrats try to figure out how to move forward.
Vice versa, are there lessons to be drawn from the races that underperformed most? Ted Strickland was buried by attack ads in a race where campaign funding was among the most unevenly distributed. He was also dragged down by having governed the state right when the world economy tanked in 2008. Sue Minter was up against a personally very popular, middle-of-the-road Republican. It's still striking how much they trailed Clinton, though. It's even more striking how Patty Judge bombed so badly in Iowa. She performed worse, by this measure, than Strickland.
There are other ways to explore whether these Democratic candidates overperformed or underperformed, though. For example, how did they fare compared with what the polls had been predicting?
OUTPERFORMING - OR UNDERPERFORMING - the polls
STATE |
DEM CANDIDATE |
GOP CANDIDATE |
DEM MARGIN |
polling MARGIN |
DIFFERENCE |
|
US SENATE |
|
|
|
|
ILLINOIS |
Duckworth |
Kirk |
14.2 |
13 |
1.2 |
COLORADO |
Bennett |
Glenn |
3.9 |
9 |
-5.1 |
NEVADA |
Masto |
Heck |
2.4 |
1 |
1.4 |
NEW HAMPSHIRE |
Hassan |
Ayotte |
0.1 |
2 |
-1.9 |
PENNSYLVANIA |
McGinty |
Toomey |
-1.7 |
4 |
-5.7 |
MISSOURI |
Kander |
Blunt |
-3.2 |
0 |
-3.2 |
WISCONSIN |
Feingold |
Johnson |
-3.4 |
4 |
-7.4 |
NORTH CAROLINA |
Ross |
Burr |
-5.8 |
-2 |
-3.8 |
FLORIDA |
Murphy |
Rubio |
-7.7 |
-6 |
-1.7 |
INDIANA |
Bayh |
Young |
-9.7 |
1 |
-10.7 |
ARIZONA |
Kirkpatrick |
McCain |
-12.3 |
-8 |
-4.3 |
GEORGIA |
Barksdale |
Isakson |
-14.2 |
-10 |
-4.2 |
OHIO |
Strickland |
Portman |
-21.4 |
-16 |
-5.4 |
IOWA |
Judge |
Grassley |
-24.5 |
-16 |
-8.5 |
|
GOVERNOR |
|
|
|
|
WEST-VIRGINIA |
Justice |
Cole |
6.8 |
5 |
1.8 |
MONTANA |
Bullock |
Gianforte |
3.8 |
n/a |
|
NORTH CAROLINA |
Cooper |
McCrory |
0.1 |
3 |
-2.9 |
NEW HAMPSHIRE |
Van Ostern |
Sununu |
-2.3 |
3 |
-5.3 |
MISSOURI |
Koster |
Greitens |
-5.9 |
5 |
-10.9 |
INDIANA |
Gregg |
Holcomb |
-6.0 |
5 |
-11.0 |
VERMONT |
Minter |
Scott |
-8.7 |
n/a |
|
Data sources: NY Times (Senate), Politico (Governor), Huffington Post Pollster (polling trendlines). Where HuffPo Pollster did not identify enough polls to define a polling trend, I've put n/a.
The overall picture here is even more depressing than above. In all but three races, the polls had foreseen a better result for the Democrat than what emerged. (At least when using HuffPo Pollster as source; in the Presidential race, Pollster identified larger leads for Clinton than e.g. RealClearPolitics.) The polls had even foreseen a sweep of the gubernatorial races, unless you count a last-minute SurveyMonkey poll which saw a narrow loss for Steve Bullock in a race that lacked enough polls to define a trendline. In the end, the Democrats won less than half of them.
The three candidates who did better than the polls had suggested were Tammy Duckworth, Catherine Cortez Masto, and Jim Justice. Very different candidates in very different races, and each of their winning margins outperformed the polls by less than two points.
The candidates whose chances were overestimated most by the polls were some of the same who still ended up outperforming Clinton by the greatest lengths: Evan Bayh, Chris Koster, John Gregg. This doubtlessly has a lot to do with just how much Trump outperformed the polls in those red states at the top of the ticket. Pollster had Trump leading by ten points in Indiana, and seven in Missouri; in both cases the real margin turned out to be 19 points. The Republican candidates for Senator and Governor outperformed the polls by roughly the same amount.
Bayh faced some strong headwinds at the end of his campaign over revelations about his lobbying work and (lack of) residence in his supposed home state. So it's interesting that, in the end, he heavily underperformed the polls, but not by more than Gregg did as gubernatorial candidate in the same state.
Beyond Bayh, Koster and Gregg, the candidate who underperformed the polls most was Patty Judge, who was at the bottom of the list in the comparison with Clinton.
I could think of one last way to try to evaluate the performance of these Democratic candidates. It's an obvious enough one: how did they do compared to the last election?
This gets a little tricky, however. The last time these Senate races were on was in 2010, an annus horribilis for Congressional Democrats. The last time the gubernatorial races were on was in 2012, a pretty good year for Democrats. So the performance of the Democratic candidates on the two lists cannot be easily compared when using this yardstick:
OUTPERFORMING - OR UNDERPERFORMING - the previous election
STATE |
DEM CANDIDATE |
GOP CANDIDATE |
DEM MARGIN |
DEM MARGIN |
DIFFERENCE |
|
US SENATE |
|
|
2010 |
|
ILLINOIS |
Duckworth |
Kirk |
14.2 |
-1.6 |
15.8 |
COLORADO |
Bennett |
Glenn |
3.9 |
-1.7 |
5.6 |
NEVADA |
Masto |
Heck |
2.4 |
5.7 |
-3.3 |
NEW HAMPSHIRE |
Hassan |
Ayotte |
0.1 |
-23.2 |
23.3 |
PENNSYLVANIA |
McGinty |
Toomey |
-1.7 |
-2.0 |
0.3 |
MISSOURI |
Kander |
Blunt |
-3.2 |
-13.4 |
10.2 |
WISCONSIN |
Feingold |
Johnson |
-3.4 |
-4.8 |
1.4 |
NORTH CAROLINA |
Ross |
Burr |
-5.8 |
-11.8 |
6.0 |
FLORIDA |
Murphy |
Rubio |
-7.7 |
3-way race |
|
INDIANA |
Bayh |
Young |
-9.7 |
-14.6 |
4.9 |
ARIZONA |
Kirkpatrick |
McCain |
-12.3 |
-24.2 |
11.9 |
GEORGIA |
Barksdale |
Isakson |
-14.2 |
-19.3 |
5.1 |
OHIO |
Strickland |
Portman |
-21.4 |
-17.5 |
-3.9 |
IOWA |
Judge |
Grassley |
-24.5 |
-31.1 |
6.6 |
|
GOVERNOR |
|
|
2012 |
|
WEST-VIRGINIA |
Justice |
Cole |
6.8 |
4.7 |
2.1 |
MONTANA |
Bullock |
Gianforte |
3.8 |
1.6 |
2.2 |
NORTH CAROLINA |
Cooper |
McCrory |
0.1 |
-11.4 |
11.5 |
NEW HAMPSHIRE |
Van Ostern |
Sununu |
-2.3 |
12.1 |
-14.4 |
MISSOURI |
Koster |
Greitens |
-5.9 |
12.3 |
-18.2 |
INDIANA |
Gregg |
Holcomb |
-6.0 |
-2.9 |
-3.1 |
VERMONT |
Minter |
Scott |
-8.7 |
20.2 |
-28.9 |
Data sources: NY Times (2016 Senate), Politico (2016 Governor), FEC (2010 Senate); Ballotpedia (2012 Governor)
In the category of small favours, this year's Senate results don't look so bad when you compare them to 2010. All but two of the candidates in (at some point) competitive-seeming Senate races this year did better than the Democrat in that race in 2010. Even the exceptions (Masto and Strickland) only had a 3-4 point worse margin, though it's worth noting that Strickland actually did worse than the unheralded Lee Fisher.
The greatest improvements were accomplished in New Hampshire, where Hassan unraveled the 23-point buffer Ayotte built up last time, when she ran against Paul Hodes; and Illinois, where Duckworth turned a narrow Kirk victory over Alexi Giannoulias in 2012 into a major defeat for him now. The power of incumbency did not help either Republican very much at all.
Other double-digit improvements over previous results were achieved by Jason Kander in Missouri, where Robin Carnahan had lost by 13 points in 2010; and Ann Kirkpatrick in Arizona, where Rodney Glassman suffered an ignominious 24-point defeat in 2010.
The Democrats running for Governor this year had a tougher benchmark to meet in the 2012 election results, and most didn't make it. The big exception is North Carolina's Roy Cooper. While doing a little worse than the polls had suggested and only winning by a hairwidth, he still proved a major upgrade over 2012 candidate Walter Dalton, who'd lost by a double-digit margin.
In Montana, now benefiting from incumbency, Steve Bullock improved his winning margin a little. In West-Virginia, Jim Justice outdid his predecessor Earl Ray Tomblin by a little, at least in terms of his winning margin. Not so in terms of his share of the vote: Tomblin had gotten 51% to his Republican opponent's 46%, while Justice won 49% to 42%. Both parties’ candidates performed worse this year as the environmentalist Mountain Party increased its share from 2.5% to 6.4%.
Major backsliding in Democratic performance took place in New England, however, where Van Ostern and Minter far underperformed the Democrats who had won the same races in 2012; as well as in Missouri, where the margin now was 18 points worse than in 2012.
So whom should national Democrats turn to for some potentially useful advice by this metric? It's a mostly different set of names than the first table, above, suggested: Duckworth, Hassan, Kirkpatrick... Jason Kander's the only name that emerges again.
One could argue that their stand-out improvements over the Democrats’ 2012 scores are somewhat less than meets the eye because the 2012 candidates in those states were particularly weak. But Hassan and Duckworth still succeeded in turning out, respectively humiliating, two incumbents, and Hassan won her race in an intensely-fought swing state.
On the flip side, are there lessons to be learnt from the weak performances of Sue Minter and Colin Van Ostern? Minter has the excuse of facing a particularly strong opponent, but she did underperform both Peter Shumlin and Hillary Clinton by truly imposing amounts. And Van Ostern underperformed by all measures. His result was considerably worse than Maggie Hassan's performance when she won the gubernatorial office four years ago; worse than the polls had suggested; and worse than Hillary Clinton's performance in the state.