As I seem to have a love of spreadsheets, here’s a Google spreadsheet that will keep track of the daily Early Voting returns by location. Counting mail ballots and absentee in-person voting, 6162 people voted yesterday. For comparison purposes, here’s how the vote broke down in each of the last three elections:
Year Mail Early Total Pct Early =========================================== 2003 17,583 76,285 374,459 25.07 2005 10,703 72,361 332,154 25.00 2007 6,844 43,420 193,945 25.92
All numbers cited in the chart above are for Harris County as a whole, so the comparison is apples to apples. Note that 2005 had a large number of non-City of Houston votes, thanks to the anti-gay marriage ballot proposition. I have a Google doc here that shows the EV totals for each day in 2003, 2005, and 2007, and which also shows that on average in odd numbered election years since 1997, the City of Houston accounts for about 77% of all ballots cast in Harris County.
Breaking that down somewhat, you get about 4745 City of Houston ballots for yesterday, assuming the historic rate holds true, and 79,719 total COH early votes from 2003, which had a Houston/Harris percentage of 79.6. A large number of the early votes in 2003 were cast on the last two days – 30,027 in-person votes, or 39% of the total, came on the Thursday and Friday of the second week. It was similar in 2005 and 2007 as well – 41% and 36%, respectively, of all in-person early votes were cast on the last two days those years.
You don’t really want to project too much from one day’s total. But if you do compare the first-day in-person totals from these last three elections to this year’s, you get the following result:
Year Early Day 1 % Day 1 Tot EV ===================================== 2003 2,977 3.90% 76,285 2005 2,451 3.39% 72,361 2007 1,681 3.87% 43,420 2009 4,089 3.70% * 110,513 *
Asterisks indicate projected totals; the percent of Day One for 2009 is the average of the other three years. I happen to believe we’ll see a higher rate of early voting overall this year than in years past – I think it’s more of a habit now after the experience of the 2008 election – and as such I don’t believe for a minute that we’ll get the 450,000 county voters, or 350,000 city voters, that this would extrapolate to if previous years’ habits held steady. But this does at least suggest the possibility to me that we’ve all underestimated the turnout level. Maybe this year will be more like 2003 than we think.
Anyway. I’ll update the spreadsheet as each day’s returns come in, and will hone my guess as to final turnout as we go. My thanks to Hector DeLeon of the County Clerk’s office for providing me with the data from previous years.
UPDATE: Day Two results are now in – another 4206 in-person ballots. 2009 continues to run ahead of 2003. Either we’re seeing a big shift in early voting behavior – more of it early in the period, with less in the last two days, or more of if as a share of the overall vote – or we’re in for a much higher turnout this year than originally thought.