There are two things I left off of the 2004 data. The first is the percentage of Delaware voters voting Democratic in the 2004 general election. As the world is well-aware, Delaware has turned a deep shade of blue, contrary to its previous reddish hue in the eighties and early nineties. The total percentage voting "D" in the 2004 general election was 53.3%; in New Castle, the percentage was 60.5%, in Kent, 42.6%, and in Sussex, 38.7%. Keep those percentages in mind.
Another chart necessary for comparison is the percentage of total vote each candidate received in the 2004 general election. That chart is here:
| Percentage of total vote | | | | |
| 2004 | Total | New Castle | Kent | Sussex |
| Kerry | 53.3% | 60.5% | 42.6% | 38.7% |
| Bush | 45.8% | 38.5% | 56.4% | 60.5% |
| Nader | 0.57% | 0.55% | 0.65% | 0.58% |
| Other | 0.33% | 0.36% | 0.29% | 0.25% |
| % population casting vote | 45.3% | 46.6% | 40.3% | 45.2% |
| % voting population casting vote | 60.6% | 62.1% | 53.8% | 60.3% |
With the foundation set, here is the 2008 data (compiled from CNN's primary data and related sites).
2008 Delaware Republican primary voter turnout (Feb. 2008):
| Feb. 2008 Republican | Total | New Castle | Kent | Sussex |
| McCain | 22,628 | 13,227 | 3,598 | 5,803 |
| Romney | 16,344 | 8,758 | 2,806 | 4,780 |
| Huckabee | 7,706 | 3,140 | 1,568 | 2,998 |
| Paul | 2,131 | 1,434 | 289 | 408 |
| Total | 48,809 | 26,559 | 8,261 | 13,989 |
| Population estimate | 871,674 | 530,481 | 154,592 | 186,749 |
| Voting population estimate | 653,756 | 397,861 | 115,944 | 140,062 |
| % population casting vote | 5.6% | 5.0% | 5.3% | 7.5% |
| % voting pop. casting vote | 7.5% | 6.7% | 7.1% | 10.0% |
2008 Democrat Primary voter turnout (Feb. 2008):
| Feb. 2008 | Total | New Castle | Kent | Sussex |
| Obama | 51,148 | 37,818 | 6,736 | 6,594 |
| Clinton | 40,760 | 26,572 | 5,534 | 8,654 |
| Biden | 2,863 | 1,758 | 408 | 670 |
| Edwards | 1,241 | 575 | 273 | 393 |
| Other | 362 | 198 | 64 | 100 |
| Total | 96,374 | 66,948 | 13,015 | 16,411 |
| Population estimate | 871,674 | 530,481 | 154,592 | 186,749 |
| Voting population estimate | 653,756 | 397,861 | 115,944 | 140,062 |
| % population casting vote | 11.1% | 12.6% | 8.4% | 8.8% |
| % voting population casting vote | 14.7% | 16.8% | 11.2% | 11.7% |
Now, this doesn't really mean too much until you look at it broken down by party, and then compare the 2004 data to it. Granted, it's tough to compare exactly because there was no meaningful Republican primary (if there even was one at all) in 2004 because Bush was running for reelection. Nevertheless, nearly 40% of the turnout from the 2004 general election came out to vote in the 2008 primary. That's a huge increase. More interestingly, the turnout of those voting Democrat was statistically increased over the 2004 election. In other words, all three counties "turned blue", so to speak. Most interestingly, however, is the turnout in February 2008 as compared to February 2004. While the population increased only approximately 6% over that time period, the increase in voter turnout represented nearly a 300% increase.
| Feb 2008 | Total | New Castle | Kent | Sussex |
| Democrat | 96,374 | 66,948 | 13,015 | 16,411 |
| Republican | 48,809 | 26,559 | 8,261 | 13,989 |
| Total | 145,183 | 93,507 | 21,276 | 30,400 |
| Population estimate | 871,674 | 530,481 | 154,592 | 186,749 |
| Voting population estimate | 653,756 | 397,861 | 115,944 | 140,062 |
| % population casting vote | 16.7% | 17.6% | 13.8% | 16.3% |
| % voting population casting vote | 22.2% | 23.5% | 18.4% | 21.7% |
| Percentage of 2004 voting numbers | 38.7% | 38.7% | 38.0% | 39.1% |
| Percentage voting Democrat | 66.4% | 71.6% | 61.2% | 54.0% |
All of this means very little, until you start looking at the population numbers. Using an estimated 1.0137% annual growth rate from July 2007 (which breaks down to a monthly growth rate of 0.061% in New Castle, 0.218% in Kent, and 0.189% in Sussex), and assuming a constant 75% voting population estimate, I derive the following chart (with registered voter data coming from the Department of Elections).
| July 2008 Population Data | Total | New Castle | Kent | Sussex |
| Population Estimate | 876,688 | 532,090 | 156,235 | 188,481 |
| Voting population estimate | 657,516 | 399,068 | 117,176 | 141,361 |
| Registered voters | 580,046 | 366,073 | 95,388 | 118,585 |
The initial question I had was whether the staggering increase in voter turnout in February would still hold true today. Fortunately for Delaware, the gubernatorial race between Markell and Carney proved worthwhile in a statistical sense, and provides support for my hypothesis that the answer is yes. While the numbers are slightly lower, it still demonstrates that voters continue to be engaged.
| September 2008 Democrat Governor Primary | Total (including absentee votes) |
| Carney | 36,112 |
| Markell | 37,849 |
| Total | 73,961 |
| Population estimate | 878,565 |
| Voting population estimate | 658,924 |
| % population casting vote | 8.4% |
| % voting population casting vote | 11.2% |
Thus, I am left to predict using this data (and my previous data calculations) to derive the turnout in November 2008. Because it is impossible to predict exactly how this election will play out, I will simply combine the data to show what the numbers could be using the 2004 percentages and the 2008 party distribution.
| November 2008 | Total | New Castle | Kent | Sussex |
| Est. Population | 880,566 | 533,392 | 157,601 | 189,914 |
| Est. Voting population | 660,425 | 400,044 | 118,201 | 142,436 |
| Using 2004 voting percentages | | | | |
| Total | 400,298 | 248,621 | 63,557 | 85,891 |
| Obama | 213,547 | 150,514 | 27,106 | 33,250 |
| McCain | 183,148 | 95,839 | 35,852 | 51,925 |
| Nader | 2,297 | 1,376 | 412 | 502 |
| Barr/Other | 1,307 | 893 | 186 | 214 |
| Using 2008 percentages voting Democrat | | | | |
| Obama | 265,722 | 178,005 | 38,879 | 46,367 |
| McCain | 130,573 | 68,130 | 24,042 | 38,665 |
| Other (1%) | 4,003 | 2,486 | 636 | 859 |
Thus, if Delaware had a similar voting turnout to 2004, Obama will likely carry New Castle County, and carry enough of Kent and Sussex to give him three electoral votes. This is not surprising, and every national poll has Delaware going blue. What I think is more interesting, however, is what happens if Obama (and the Democratic party) is successful in continuing to mobilize this "new" base (as evidenced by the incredible increase in voter turnout in February). Applying those numbers, as the chart indicates, Obama would carry each county, and carry the state with 66% of the vote.
The real question, in my mind, is not whether the February increase in voter turnout will carry over (since the September data results indicates it will), it is a question by how much. If the voter turnout increases from 60% to say, 70%, the numbers statistically favor a Democrat victory.
Now, this is just Delaware. Third party candidates have a statistically insignificant impact, and the population distribution of the state favors the Democrat party. What I find fascinating in this election is whether other states (and I suppose those that are "toss-ups") have similar percentage increases in voter turnout. Assuming for the sake of argument that they do, then that would seem to suggest that the percentages of voters voting Democrat in the various pockets of other states would also increase. My theory concerns that percentage increase in these pockets over those areas that experience a less dramatic increase in voter turnout (a theory that would obviously require analysis of every county in every state, which is simply beyond the scope of my three-part series). In the end, it becomes a simple numbers game. For those states where Democrats come out strong (such as in Virginia, Colorado, and other "toss up" states), the mobilization certainly favors the Democrats.
At this point, the only remaining question, which we will find out in a couple more months, is whether the voter turnout stays the same as in November 2004, or meets or exceeds that of February 2008. For John McCain, the answer to that question may very well depend on a mobilized base, which, in this year of it not being good to be the party in power, probably is a long-shot. Thus, while the election polls are running their regular polling data, and they should, I simply look at this data and conclude that the numbers are not taking into consideration the higher-than-predicted voter turnout. This x-factor certainly will be something all the reporters will be examining more heavily on the day and night of the election.
Tags:
No comments:
Post a Comment