Wednesday, November 22, 2006
More on Florida CD 13 ballot design
The Herald Tribune has this story on the ballot design hypothesis for the massive undervote in Sarasota County's 13th Congressional race. Here is a snip from the story:
Here is a link to a pdf file that has some of the ballots (I received this from MIT colleague Ted Selker).
On election day, voters in Sarasota and Charlotte counties went through much the same process.
They used the same type of electronic voting machines, made by the same company. In certain precincts, they voted in the same major races -- U.S. Senate, governor, U.S. House and state attorney general.
But there were subtle differences in design of the ballot from one county to the next. And as the votes were counted, those differences showed up in unexpectedly large undervotes in the U.S. House and attorney general's races, according to ballot design experts.
Sarasota recorded a massive undervote in the House District 13 race and a typical undervote in the attorney general's race. Charlotte County had a typical undervote in the District 13 race and massive undervote in the attorney general's race.
Experts say the undervotes were likely caused by badly designed ballots that led people to accidentally skip over one race as they scrolled through their computer screens.
Here is a link to a pdf file that has some of the ballots (I received this from MIT colleague Ted Selker).
My response to Bill Bradbury's Washington Post OpEd
My response to Secretary of State Bill Bradbury's recent Washington Post oped:
Secretary Bradbury is rightly proud of the tradition of high turnout and clean elections in Oregon. We have to be careful, however, in how much we can generalize from the Oregon experience to elections nationwide.
Oregon has always been a participative state. In the three midterm elections prior to voting by mail, Oregon turnout averaged 72%. Turnout in the three midterm elections after voting by mail averaged 66%. There is simply no evidence that voting by mail increased overall turnout in the state--and it is certainly incorrect to point to our turnout numbers in 2006 (69%) as prima facie evidence of the impact of this voting system.
Furthermore, Oregon's turnout rate is not easily compared to other states due to our very clean registration rolls. The way Oregon maintains its rolls is a positive side effect of voting by mail (although not a reason that it was originally adopted). Essentially, Oregon relies on the US Postal Service to help maintain the rolls. Ballots are not forwarded, so if you move between elections and fail to notify the elections office, you are sent a postcard reminder to update your registration record. This keeps a lot of deadwood off the rolls (it also means that the registration "hurdle" is higher in Oregon).
But most important for our purposes here, it means that the "denominator" in the typical turnout calculation (ballots / registered voters) is smaller in Oregon than in other states with less clean rolls. If you calculate turnout by (ballots/eligible voters), Oregon still looks good, but it appears far more typical to other homogeneous, relatively affluent states with low commute times.
It is true that voting by mail is cheaper than "mixed" systems (systems with high levels of both precinct and absentee balloting)--though no one has yet compared costs to systems with high levels of in-person early voting and precinct place voting. And he's also right that Oregon citizens endorse continued use of voting by mail.
There is a national movement underway, led by Senator Ron Wyden, Secretary Bill Bradbury, and other Oregonians to implement voting by mail nationwide. Like with any innovation, voting by mail has both advantages and disadvantages. What's most important is that before we move wholesale to another major electoral reform, we understand its impact as fully as possible.
Secretary Bradbury is rightly proud of the tradition of high turnout and clean elections in Oregon. We have to be careful, however, in how much we can generalize from the Oregon experience to elections nationwide.
Oregon has always been a participative state. In the three midterm elections prior to voting by mail, Oregon turnout averaged 72%. Turnout in the three midterm elections after voting by mail averaged 66%. There is simply no evidence that voting by mail increased overall turnout in the state--and it is certainly incorrect to point to our turnout numbers in 2006 (69%) as prima facie evidence of the impact of this voting system.
Furthermore, Oregon's turnout rate is not easily compared to other states due to our very clean registration rolls. The way Oregon maintains its rolls is a positive side effect of voting by mail (although not a reason that it was originally adopted). Essentially, Oregon relies on the US Postal Service to help maintain the rolls. Ballots are not forwarded, so if you move between elections and fail to notify the elections office, you are sent a postcard reminder to update your registration record. This keeps a lot of deadwood off the rolls (it also means that the registration "hurdle" is higher in Oregon).
But most important for our purposes here, it means that the "denominator" in the typical turnout calculation (ballots / registered voters) is smaller in Oregon than in other states with less clean rolls. If you calculate turnout by (ballots/eligible voters), Oregon still looks good, but it appears far more typical to other homogeneous, relatively affluent states with low commute times.
It is true that voting by mail is cheaper than "mixed" systems (systems with high levels of both precinct and absentee balloting)--though no one has yet compared costs to systems with high levels of in-person early voting and precinct place voting. And he's also right that Oregon citizens endorse continued use of voting by mail.
There is a national movement underway, led by Senator Ron Wyden, Secretary Bill Bradbury, and other Oregonians to implement voting by mail nationwide. Like with any innovation, voting by mail has both advantages and disadvantages. What's most important is that before we move wholesale to another major electoral reform, we understand its impact as fully as possible.
Review of election problems in Boston
An old friend and colleague of mine, David King, is heading up a review of Boston's Election Department, the The Boston Globe reports. Boston had problems with its machines; it ran out of ballots at some locations; and there were issues with disabled voter access.
Dave's a good egg, a real pro, and you can be sure this will be a comprehensive and non-partisan review. More info will be forthcoming as I can share it.
Dave's a good egg, a real pro, and you can be sure this will be a comprehensive and non-partisan review. More info will be forthcoming as I can share it.
Monday, November 20, 2006
Early Voting and the 2008 Presidential Primaries
It's a standard line for most political commentators: the 2008 presidential election season has already begun. Some candidates have taken their names off of the list of contenders, while others have formed exploratory committees.
At electionupdates, we've also begun to think about the 2008 election as well, and how changes in election machinery and election administration may alter the dynamics of the 2008 contest.
Here's our first 2008 headline: Earlyvoting holds the possibility of fundamentally altering the 2008 primary calendar.
The presidential primary calendar traditionally stars with the Iowa caucuses, followed a week later by New Hampshire. But what voters in other states who may begin casting their ballots early? California, for instance, is currently scheduled to hold its primary in early March (with proposals under consideration to move it even earlier).
When does the California primary really begin? Absentee ballots arrive as early as a month prior to the date of the election. If a similarly long window is allowed in 2008, then 10% or more of the primary ballots will be cast before the New Hampshire primary!
And if California moves its primary to February, as is being proposed, the first ballots could be case in mid January.
The impact on campaign strategy is obvious: you have to pay attention to the early voting calendar as well as the election day calendar. Only well-funded campaigns are going to be able to maintain a multi-state mobilization effort that early in the season. And the likelihood of a dark horse emerging from Iowa and New Hampshire is reduced even further.
Read about the early voting calendar at earlyvoting.net
At electionupdates, we've also begun to think about the 2008 election as well, and how changes in election machinery and election administration may alter the dynamics of the 2008 contest.
Here's our first 2008 headline: Earlyvoting holds the possibility of fundamentally altering the 2008 primary calendar.
The presidential primary calendar traditionally stars with the Iowa caucuses, followed a week later by New Hampshire. But what voters in other states who may begin casting their ballots early? California, for instance, is currently scheduled to hold its primary in early March (with proposals under consideration to move it even earlier).
When does the California primary really begin? Absentee ballots arrive as early as a month prior to the date of the election. If a similarly long window is allowed in 2008, then 10% or more of the primary ballots will be cast before the New Hampshire primary!
And if California moves its primary to February, as is being proposed, the first ballots could be case in mid January.
The impact on campaign strategy is obvious: you have to pay attention to the early voting calendar as well as the election day calendar. Only well-funded campaigns are going to be able to maintain a multi-state mobilization effort that early in the season. And the likelihood of a dark horse emerging from Iowa and New Hampshire is reduced even further.
Read about the early voting calendar at earlyvoting.net
