Shortly after the Tuesday primary election results were announced, a table of Democratic primary voter turnout since 1990 starting popping up, showing that turnout in 2018 was the largest in history.
While it was nice to see that information, it immediately made me wonder what primary voter turnout might mean for the general election, while also wondering what turnout was like on the Republican side. My Google Fu skills are not great, and I couldn’t find more comprehensive data out there. So I decided to create a table myself.
What follows is a table showing (by mid-term election years): 1) primary voter turnout for Republicans and Democrats in the Gubernatorial Primary; 2) general election votes for the Republican and Democratic candidate; and 3) the percentage increase from primary turnout to general election votes.
Comparison of Gubernatorial Primary and General Election Votes by Party 1990-2014
|
(R) Primary Votes |
(R) General Election Votes |
% Increase |
(D) primary Votes |
(D) General Election Votes |
% Increase
|
2018 |
1,618,013
|
? |
? |
1,508,984
|
? |
? |
2014 |
949,144
|
2,865,343
|
201%
|
837,796
|
2,801,198
|
234%
|
2010 |
1,294,438
|
2,619,335
|
102%
|
871,335
|
2,557,785
|
194%
|
2006 |
985,986
|
2,519,845
|
155%
|
857,814
|
2,178,289
|
154%
|
2002 |
966,875
|
2,856,845
|
195%
|
1,357,017
|
2,201,427
|
62%
|
1998 |
998,566
|
2,191,105
|
119%
|
1,244,044
|
1,773,054
|
43%
|
1994 |
901,237
|
2,071,068
|
130%
|
836,414
|
2,135,008
|
155%
|
1990 |
668,181
|
1,535,068
|
130%
|
1,074,056
|
1,995,206
|
86%
|
Average |
|
|
147%
|
|
|
132%
|
Some explanation of these numbers. First, primary votes includes votes for all candidates running in the respective party primaries. It is not limited to the votes of the party primary winner. In some cases, a candidate ran unopposed (e.g. in 1998, both the Democratic and Republican candidates ran unopposed in the party primaries), while this year there were 7 candidates running in the Democratic primary. Second, the general election vote count represents the total votes received by the party nominee in the general election. Votes for third party or independent candidates are not considered. So we can look at 2014, for example, and quickly see that Republican candidates received a total of 949,144 votes in the primary. Rick Scott, in turn, received 2,865,343 votes in the general election, an increase of 201% over the primary turnout.
As you can see, both Republican and Democratic primary voter turnout in 2018 was the highest in at least the last 28 years. What I wanted to know is, what can we expect to happen in November for the general election based on what happened this week?
Over the last 7 gubernatorial elections, votes for the Republican candidate increased about 147% from the total Republican primary voter turnout. In turn, votes for the Democratic candidates increased about 132%. On the surface, this runs counter to the stereotype that Democrats as a whole don’t vote as consistently as Republicans. In other words, you’d expect that if Democrats more typically only bothered to vote in the general election, then Democratic candidates would typically see a larger increase in votes between the primary and the general election. But since 1990, it is Republicans who have generally enjoyed the larger turnout increase.
If those average increases held true this year, voter turnout would shatter all previous records, and DeSantis would win with 3,996,492 votes (53%) to Gillum’s 3,500,842 (47%). However, previous elections show a pretty consistent trend that when primary turnout is high, then the corresponding increase in the general election is relatively smaller. In the three primaries where Democrats received over one million votes, the increase in votes for the Democratic candidate averaged about 63%. Republicans, in contrast, have only broken one million primary votes once, in 2010, and the general election vote increased by about 100% that year.
I don’t have the ability to draw any firm conclusions about how general election turnout will increase this year compared to the primary vote totals. Of course, there are extenuating factors, such as the quality (or lack thereof) of the candidates, local issues such as the ongoing Red Tide that is affecting parts of the Florida coast, and national issues such as Trump’s popularity. However, if I had to guess, I suspect turnout will end up roughly double for each party, with more than enough uncertainty between the parties to expect that Gillum can win.
On a related note, voter registration numbers over the last 28 years show that Republicans have steadily gained on Democrats. In 1990, registered Democrats exceeded registered Republicans by about 700,000. Since then, Republicans have grown by about 2 million, while Democrats have increased by about 1.7 million. Today, the difference less than 350,000. However, it’s still the case that Democrats have more room to grow in the general election.
At the same time, the fastest growing group is “Other”, which has grown from about 433,000 to 3.5 million — an increase of about 700%. I don’t know enough about the Florida Electorate to say what political leaning this group may have as a whole. But considering that Democrats haven’t won a gubernatorial election in over 20 years, this group is probably even split at best.
Year |
Republican Party of Florida |
Florida Democratic Party |
Other |
Total |
2016 |
4,575,277 |
4,905,705 |
3,478,203 |
12,959,185 |
2015 |
4,193,586 |
4,520,265 |
3,229,616 |
11,943,467 |
2014 |
4,182,775 |
4,627,737 |
3,175,916 |
11,986,428 |
2013 |
4,133,350 |
4,621,791 |
2,962,605 |
11,717,746 |
2012 |
4,263,587 |
4,821,859 |
2,953,125 |
12,038,571 |
2011 |
4,061,224 |
4,552,483 |
2,622,454 |
11,236,161 |
2010 |
4,042,393 |
4,611,335 |
2,562,010 |
11,215,738 |
2009 |
3,967,472 |
4,637,354 |
2,459,541 |
11,064,367 |
2008 |
4,106,743 |
4,800,890 |
2,504,290 |
11,411,923 |
2007 |
3,826,836 |
4,138,604 |
2,241,161 |
10,206,601 |
2006 |
3,920,201 |
4,196,608 |
2,268,797 |
10,385,606 |
2005 |
3,954,304 |
4,276,512 |
2,241,102 |
10,471,918 |
2004 |
3,954,492 |
4,322,376 |
2,199,569 |
10,476,437 |
2003 |
3,577,179 |
3,880,342 |
1,808,963 |
9,266,484 |
2002 |
3,610,992 |
3,956,694 |
1,756,873 |
9,324,559 |
2000 |
3,474,438 |
3,853,524 |
1,552,434 |
8,880,396 |
1998 |
3,327,207 |
3,731,367 |
1,268,133 |
8,326,707 |
1996 |
3,344,036 |
3,774,809 |
1,077,812 |
8,196,657 |
1994 |
2,747,074 |
3,245,518 |
567,006 |
6,559,598 |
1992 |
2,672,968 |
3,318,565 |
550,292 |
6,541,825 |
1992 |
2,672,968 |
3,318,565 |
550,292 |
6,541,825 |
1990 |
2,448,488 |
3,149,747 |
432,926 |
6,031,161 |
This is my first post in a few years, so please be kind. Having said that, I’d love some constructive feedback on how to analyze this information in a more accurate and useful way.