With just over 24 hours until the first polls close, more than 35 million Americans have already voted in the 2018 midterm elections — a 176 percent increase over the 20 million people who voted at this point in 2014. While early vote data cannot predict election day turnout, there are key insights that we can draw from the electorate that has already turned out, including that the electorate is on course to be younger and more diverse than any midterm in history. For more early vote data, visit TargetEarly, a one-stop interactive dashboard of early vote data for 42 states.
The Early Electorate by Voter Propensity
Of those who have already voted, more than 1.5 million are new registrants who have never voted before and 6 million are infrequent voters. These two groups are seeing a .92% and 4.17% growth in total vote share, respectively, over the 2014 midterms. By comparison, the share of early vote among voters most-likely to turn out (so-called “super voters”) are down a significant 8.35 percent. The number of voters who have never voted or are infrequent votes has more than doubled compared to 2014.
Voter Score | 2018 EAV | 2014 Compare (days out) | 2018 Share | 2014 Compare Share | Share Difference | 2014 Compare % Change |
---|---|---|---|---|---|---|
Super Voter | 12,608,825 | 8,871,586 | 36.24% | 44.59% | -8.35% | 42.13 % |
Frequent Voter | 14,532,795 | 7,685,589 | 41.77% | 38.63% | 3.14% | 89.09% |
Infrequent Voter | 6,022,651 | 2,613,908 | 17.31% | 13.14% | 4.17% | 130.41% |
Never Voted | 1,581,972 | 722,562 | 4.55% | 3.63% | 0.92% | 118.94% |
Unknown | 43,321 | 599 | 0.12% | 0.00% | 0.12% | 7132.22% |
The Early Electorate by Gender
Women are voting nationally at about the same levels as they did in 2014, which means they continue to dominate turnout by a roughly 54 percent to 45 percent margin. Already, six million more women have already voted this cycle compared to the same time in 2014; and three million more women have voted than men.
Though early voting levels for female voters may seem static nationwide compared to 2014, women have substantially increased their early vote turnout in key states, including Tennessee (+280 percent), Georgia (+125 percent), Pennsylvania (+120 percent), Texas (+152 percent), and Nevada (+113 percent).
Gender | 2018 EAV | 2014 Compare (days out) | 2018 Share | 2014 Compare Share | Share Difference | 2014 Compare % Change |
---|---|---|---|---|---|---|
Female | 18,901,236 | 10,766,319 | 53.61% | 53.76% | -0.16% | 75.56% |
Male | 15,882,844 | 9,203,516 | 45.05% | 45.96% | -0.91% | 72.57% |
Unknown | 475,740 | 56,013 | 1.35% | 0.28% | 1.07% | 749.34% |
The Early Electorate by Age
Younger voters are far outpacing older voters in early vote turnout compared to 2014, underscoring the increased impact youth voters are poised to have on the election. Voters aged 18–29 have increased their total early vote share by 3.21 percent compared to 2014; voters aged 30–39 have increased by 2.91 percent. Those two groups have nearly tripled their early vote rates compared to 2014. Older voters, on the other hand, have decreased their share of early vote turnout, by 2.57 percent for voters aged 50–64 and a significant 4.9 percent drop for voters 65 and older, and are lagging behind in their total early vote rates compared to 2014.
Age | 2018 EAV | 2014 Compare (days out) | 2018 Share | 2014 Compare Share | Share Difference | 2014 Compare % Change |
---|---|---|---|---|---|---|
18-29 | 2,965,695 | 1,040,732 | 8.41% | 5.20% | 3.21% | 184.96% |
30-39 | 3,429,612 | 1,364,605 | 9.73% | 6.81% | 2.91% | 151.33% |
40-49 | 4,309,653 | 2,150,243 | 12.22% | 10.74% | 1.49% | 100.43% |
50-64 | 10,563,972 | 6,514,887 | 29.96% | 32.53% | -2.57% | 62.15% |
65+ | 13,960,029 | 8.908,954 | 39.59% | 44.49% | -4.90% | 56.70% |
Unknown | 30,859 | 46,427 | 0.09% | 0.23% | -0.14% | -33.53% |
The Early Electorate by Race
Asian and Hispanic voters have more than doubled their early vote turnout compared to 2014, and African-American early vote turnout is up 77 percent. The share of early vote turnout among all racial demographics is largely the same compared to 2014, except Caucasian voters, who are 2.32 percent less of the early vote share compared to 2014.
Race | 2018 EAV | 2014 Compare (days out) | 2018 Share | 2014 Compare Share | Share Difference | 2014 Compare % Change |
---|---|---|---|---|---|---|
African-American | 3,089,523 | 1,748,086 | 8.76% | 8.73% | 0.03% | 76.74% |
Asian | 662,876 | 314,384 | 1.88% | 1.57% | 0.31% | 110.85% |
Caucasian | 28,033,018 | 16,385,386 | 79.50% | 81.82% | -2.32% | 71.09% |
Hispanic | 2,534,004 | 1,139,867 | 7.19% | 5.69% | 1.49% | 122.31% |
Other | 192,408 | 125,385 | 0.55% | 0.63% | -0.08% | 53.45% |
Uncoded | 747,991 | 312,740 | 2.12% | 1.56% | 0.56% | 139.17% |
Note on Polling
As was shown in 2016, a poll is a snapshot in time, not a prediction of future behavior. Also, a poll is only as good as it’s methodology. Every cycle, pollsters do their best to predict the likely turnout of an election. A key part of that is basing the next election on turnout in the previous comparable cycle. The challenge for pollsters is that turnout this cycle is unlike anything we’ve ever seen in a midterm. That means many pollsters — even some of the best in the field — are underweighting the very voters who are likely to shape this election.
Because the early vote totals will make up a larger percentage of the overall votes cast than in past elections, the patterns of results tomorrow might look different than what pundits and voters are used to. As voters watch the returns tomorrow night, we expect to see tighter races at the outset due to early vote, as opposed to when Democrats typically surge in early vote returns and hang on as election night progresses. So regardless of who you support or who wins, it’s going to be a very late night — and we’re certainly in for some, if not many, surprises.