The Real Lesson from Last Week’s Two Special Elections for Congress

There has been no shortage of effort by pundits and big data analysts to try to draw conclusions on whether the results of the two special elections for open House seats in Georgia and and South Carolina last week meant that Democrats or Republicans had (choose one) underperformed or overperformed, as compared to the recent district votes for President and Congress in 2016. Similar analyses followed the special elections in Kansas and Montana earlier.  

In all four cases, new Trump administration Cabinet  members who had won their district races comfortably in 2016 were replaced by Republicans who won the open seat races far less comfortably. In 3 of the 4 races, the margin for the winning Republican in the special election was narrower than Trump’s margin of victory in the district in the Presidential race last year (Georgia 6 the exception -- Trump won by a smaller percentage margin than Karen Handel).

It is highly likely, however,  that if the four new Cabinet members -- Tom Price, Mike Pompeo, Ryan Zinke and Mick Mulvaney -- had stayed in the House and would run again in 2018, they all would win easily.  In essence, special elections  are a lot different than races where incumbents are running for re-election in regular cycles, especially from generally safe districts.

Special elections are open seat races, meaning there is no incumbent.  Normally, they are held on  a day when this race is the only contested one. Turnout is usually far lower than the turnout in a normal midterm, much less a presidential year.  In the two contests last week, in districts with the same approximate population, 260,000 votes were cast for the candidates in Georgia and 87,000 for the two candidates in South Carolina.  The difference is accounted for by the amount of fundraising and media attention lavished on the Georgia, but not on the South Carolina race. Each race however wound up with a margin of victory of between 3% and 4%.

In regular election cycles, there is a big advantage to incumbency. When House seats turn over, the percentage of open seats that shift between the parties is usually far higher than the percentage of seats that turn over among the incumbents running for re-election.  If you were running the Democratic Congressional Campaign Committee for 2018, a district where the incumbent Republican is retiring and which provided a 55% to 45% margin in the last cycle, would be a far better target than a seat in which the incumbent Republican is running for re-election and also won by that same margin last time around. 

The major impact of the races last week for the GOP, particularly the closely followed Georgia election, is that it may encourage more Republicans who may have thought of retiring to stick around (They told potential candidates that the world is not ending, yet), and may slightly discourage some Democrats from thinking 2018 is a sure thing to win a Republican-held seat, damaging the party’s candidate recruitment efforts.

Why would the Georgia race have drawn three times as many voters as the South Carolina race? The weather on election day may have been worse in Georgia, where about half the votes were cast before the date of the special election.  Clearly, the media ignored the South Carolina race, assuming it was safe for the Republicans, but thought Georgia 6 would be the Democrats’ breakthrough.  The enormous campaign spending on the race on both sides, and the constant TV and radio ads, door knocking, and phone banking, seems to have fired up Democrats and NeverTrump Republicans to vote for Ossoff, but also encouraged many Republicans to show up to protect the seat for their party, particularly on Election Day. 

Winning the seat became the biggest election battle since Trump’s victory in November, and the first real chance for the left and Democrats and their media partners to slay the dragon.  The enormous amount of San Francisco and Hollywood campaign money spent in the district, made it easy to produce some clever ads for Handel mocking the out of state deluge, which may have encouraged local Republicans to show up and beat back the perceived California-funded challenge.

In South Carolina, Republicans seemed to feel the party’s control of the seat was not threatened, and stayed home, and Democrats failed to take advantage of a large African American population (near 30%) in the district to build their turnout. If $50 million has been showered on this district, turnout would likely have been far higher on both sides, maybe more so for the Republicans.

In any case, what matter is victories, and in special elections, they are not assured, even in districts which normally are strong for one party or another.  Turnout is a crapshoot, and there is no incumbency advantage.  Trump’s approval scores remain in the 40% range, but Republicans are winning open seat House contests narrowly, with 51% to 53%  of the total vote. There is simply no way to extrapolate from these four contests to 435 races 16 and a half months from now. 

Tell me how many Republicans are retiring and where their seats are, and it will be a little easier to predict the party’s chances of holding the House.  Even more important is how much legislation is passed by Congress, how the public reacts to the legislative achievements, and how far afield the special counsel chooses to go to damage the President. In any case  the percentage margin in a special election in the spring of 2017 will not matter at all.

There has been no shortage of effort by pundits and big data analysts to try to draw conclusions on whether the results of the two special elections for open House seats in Georgia and and South Carolina last week meant that Democrats or Republicans had (choose one) underperformed or overperformed, as compared to the recent district votes for President and Congress in 2016. Similar analyses followed the special elections in Kansas and Montana earlier.  

In all four cases, new Trump administration Cabinet  members who had won their district races comfortably in 2016 were replaced by Republicans who won the open seat races far less comfortably. In 3 of the 4 races, the margin for the winning Republican in the special election was narrower than Trump’s margin of victory in the district in the Presidential race last year (Georgia 6 the exception -- Trump won by a smaller percentage margin than Karen Handel).

It is highly likely, however,  that if the four new Cabinet members -- Tom Price, Mike Pompeo, Ryan Zinke and Mick Mulvaney -- had stayed in the House and would run again in 2018, they all would win easily.  In essence, special elections  are a lot different than races where incumbents are running for re-election in regular cycles, especially from generally safe districts.

Special elections are open seat races, meaning there is no incumbent.  Normally, they are held on  a day when this race is the only contested one. Turnout is usually far lower than the turnout in a normal midterm, much less a presidential year.  In the two contests last week, in districts with the same approximate population, 260,000 votes were cast for the candidates in Georgia and 87,000 for the two candidates in South Carolina.  The difference is accounted for by the amount of fundraising and media attention lavished on the Georgia, but not on the South Carolina race. Each race however wound up with a margin of victory of between 3% and 4%.

In regular election cycles, there is a big advantage to incumbency. When House seats turn over, the percentage of open seats that shift between the parties is usually far higher than the percentage of seats that turn over among the incumbents running for re-election.  If you were running the Democratic Congressional Campaign Committee for 2018, a district where the incumbent Republican is retiring and which provided a 55% to 45% margin in the last cycle, would be a far better target than a seat in which the incumbent Republican is running for re-election and also won by that same margin last time around. 

The major impact of the races last week for the GOP, particularly the closely followed Georgia election, is that it may encourage more Republicans who may have thought of retiring to stick around (They told potential candidates that the world is not ending, yet), and may slightly discourage some Democrats from thinking 2018 is a sure thing to win a Republican-held seat, damaging the party’s candidate recruitment efforts.

Why would the Georgia race have drawn three times as many voters as the South Carolina race? The weather on election day may have been worse in Georgia, where about half the votes were cast before the date of the special election.  Clearly, the media ignored the South Carolina race, assuming it was safe for the Republicans, but thought Georgia 6 would be the Democrats’ breakthrough.  The enormous campaign spending on the race on both sides, and the constant TV and radio ads, door knocking, and phone banking, seems to have fired up Democrats and NeverTrump Republicans to vote for Ossoff, but also encouraged many Republicans to show up to protect the seat for their party, particularly on Election Day. 

Winning the seat became the biggest election battle since Trump’s victory in November, and the first real chance for the left and Democrats and their media partners to slay the dragon.  The enormous amount of San Francisco and Hollywood campaign money spent in the district, made it easy to produce some clever ads for Handel mocking the out of state deluge, which may have encouraged local Republicans to show up and beat back the perceived California-funded challenge.

In South Carolina, Republicans seemed to feel the party’s control of the seat was not threatened, and stayed home, and Democrats failed to take advantage of a large African American population (near 30%) in the district to build their turnout. If $50 million has been showered on this district, turnout would likely have been far higher on both sides, maybe more so for the Republicans.

In any case, what matter is victories, and in special elections, they are not assured, even in districts which normally are strong for one party or another.  Turnout is a crapshoot, and there is no incumbency advantage.  Trump’s approval scores remain in the 40% range, but Republicans are winning open seat House contests narrowly, with 51% to 53%  of the total vote. There is simply no way to extrapolate from these four contests to 435 races 16 and a half months from now. 

Tell me how many Republicans are retiring and where their seats are, and it will be a little easier to predict the party’s chances of holding the House.  Even more important is how much legislation is passed by Congress, how the public reacts to the legislative achievements, and how far afield the special counsel chooses to go to damage the President. In any case  the percentage margin in a special election in the spring of 2017 will not matter at all.