Tuesday night, Democrats won two legislative special elections in Trump territory, flipping the seats from Republican control. The results were stunning and dramatic. In the case of New York’s 9th Assembly District, the margin shifted an astounding 39 points from last year’s presidential election!
Less noticed were two special elections the same night where party control did not change hands. Yet these, too, are part of the story. In both cases, the margin shifted just around 1 point worse compared to Hillary Clinton’s margin.
What difference does that make? Not much, until we look at the big picture.
Let’s travel back in time a bit. In December 2015, Democrat Tonya Anderson lost a special election for the 43rd District in Georgia’s state Senate by a painfully slim 84-vote margin. Here at Daily Kos Elections, we reported that “the special election gods taketh away.”
On the other hand, just last week, Democrats lost another special election for a Georgia state Senate seat, this time by a much larger 14 percentage points. But what was our headline? Republicans “shouldn’t be pleased.”
Why the difference in interpretation? Anderson was expected to win with ease. The seat was overwhelmingly Democratic; Barack Obama had won it by a whopping 44 points in 2012. Indeed, in a 2016 rematch, Anderson won the seat with a massive 70-30 advantage. What happened in 2015? Nobody really knows, but our best guess was Democrats just didn’t bother to show up.
Last week’s election, on the other hand, came in a district Obama lost by a gigantic 36-point margin, while Clinton lost it by 14 points. In previous years, we might have expected a Democrat to lose a special election by a massive amount, very likely worse than Clinton and probably even worse than Obama. But instead, the margin was the same as Clinton’s and much better than Obama’s.
Democrats are showing up—and in a way they haven’t in years. Below, we’ll show you just what’s going on.
The graph below compares Democratic margins in special elections that have taken place since Trump won to the results in those same districts in last year’s presidential election. Points above the diagonal line mean the Democrat’s margin in the special election was better than Clinton’s margin in that district. That is, of course, what we want to see. And that is, for the most part, what we do see: Democrats running in 2017 special elections are performing about the same or better than Clinton.
Below, you can see a blank version of this same graph with sectors labeled. The top half shows Democratic victories in special elections; the bottom, Republican. The territory on the right side of the graph shows districts won by the Democratic presidential candidate. On the left, they were won by the Republican.
Looking at the first graph, then, in Clinton territory, Democrats have denied Republicans any victories at all in 2017—the bottom-right corner is completely empty. And only two elections (one not shown) even fall into the “Republican moral victory” zone, just below the diagonal line in Clinton territory. Democrats have had a handful of moral victories, which, although they can be frustrating, are part of a pattern showing Democratic strength this year.
Now consider what the norm has been over the past few years. The graph below shows all the data we currently have available for special elections that took place from 2014 through 2016, compared to the 2016 presidential results:
This is clearly a very different-looking graph, with the vast majority of elections (39) below the diagonal line instead of above it (13). In other words, for the last few years, Democratic candidates in special elections have typically done about the same or worse than Clinton. Not only that, but many of the points are not just below the diagonal line, but far, far below it. Republicans were winning seats in districts that Clinton would go on to win by 10, 20, or 30 points, and even more.
Democrats? Forget about winning seats deep in Trump country. One solitary special election sits tucked away close to the origin in the upper left quadrant, barely into hostile territory. Not a fan of moral victories? Well, for the past few years, you were in luck! There were hardly any to be had for Democrats in special elections.
That is what we had come to expect: Democrats don’t tend to show up for special elections.
But that is not the case this year. This year is different.
Another way to look at this is with a histogram:
The distribution of margin changes is very different this year. In most cases, they are positive (showing the Democrat outperforming Clinton). In the recent past, this was not the case.
Note that elections with a substantial showing by third parties, major write-in candidates, jungle primaries, or candidates who switched parties are not included in this analysis. For the same analysis using 2012 Obama data, see this 2017 graph, this 2016, 2015, and 2014 graph, and this histogram.