Contrary to all reason, the Colorado Secretary of State, Republican Mike Coffman, set the transition date from county based voter registration data bases to a statewide voter registration database for late December, between two critical dates for certifying voter lists for the caucuses, rather than after January 5, 2008, when the voter lists would have already been certified, or far earlier in the year, when the lists could have been corrected without putting pressure on IT professionals to fix transition problems swiftly. I was one of many people who personally warned election officials that it was likely that there would be problems in this transition.
There were. The chickens have now come home to roost. The party affiliation of voters in the voter registration lists in the statewide database has been corrupted and the data are inaccurate. Voters are not currently in a position to confirm that they are eligible to participate in a political party caucus or confirm that their records have them listed in the precinct where they think they will caucus.
Denver's Particular Problems
Meanwhile, in other breaking news today, Denver's Clerk and Recorder has formally announced her plans to conduct Denver's election in November of 2008, in the wake of Colorado's Secretary of State Mike Coffman's apparently erroneous decision to decertify Denver's voting machines. The plan is as follows:
Denver’s voting model will consist of the following elements, based upon the committee recommendations:
Where will people vote?
o At combined precinct polling places
- Denver has 426 precincts that would be combined into approximately 175 polling locations
o At early voting sites
- Sites would be open for 10 days prior to the August 12, 2008 primary election
- Sites would be open for 15 days prior to the November 4, 2008 general election
o Through mail-in ballots
- Ballots will begin to be mailed 32 days prior to election day
- Mail-in voters may opt to be put on a permanent or election-specific mail ballot list
How will people vote?
o Paper ballots
o Direct Recording Electronic (DRE) touch screen voting machines will be available at each voting location for use by the disabled community, to be in compliance with the federal Help America Vote Act
How will votes be counted?
o Ballots will be counted at a central location on Sequoia 400c central count ballot scanners OR
o Ballots will be counted at the polling place on Sequoia Insight precinct scanners.
The biggest real problems with the clerk and recorder's plan for Denver involve the "where to vote" question and are two-fold.
First, after a disastrous experience moving from traditional local precinct voting to "vote centers" where anyone can vote at any of a number centralized locations in the City, following Larimer County's successful transition to that model (Denver vote center disaster occurred mostly as a result of faulty voter registration confirmation software), the election day system has moved to a worst of both worlds model.
Unlike the traditional precinct voting locations, the new "super-precincts" will often be at places unfamiliar to long time voters. This is driven by the Americans with Disabilities Act. Ironically, rather than helping disabled voters, who are disproportionately elderly, it will hurt them, by forcing many to abandon the familiar voting location patterns they have followed for several decades in many cases.
Lack of familiarity with your voting location is a high stakes matter this year, because unlike elections run under the "voting center" concept, which some voters may incorrectly believe is still in force in 2008, this year with super-precincts if you don't go to the proper location you don't get to vote.
The new super-precincts will also be confusing places on election day. Typically on a Presidential election day large numbers of people only dimly familiar with how voting works as a practical matter because they only vote every four years will throng out the doors of each of the super-precincts. Out of sight inside will be several different precinct voting lines, and in order to vote, voters will need to check in at the right line. Since few voters, especially those who vote infrequently, know their precinct number, the room for confusion and getting in the right sign up location is great.
Thus, super-precincts are a problem waiting to happen, and furthermore, one whose precise difficulties are unlikely to be apparent until election day when it is attempted for the first time in Denver.
Second, while the clerk and recorder's plan calls for voting both on election day and by mail-in ballot, Colorado law does not currently allow for a full fledged hybrid election in that form.
You must either affirmatively request a mail-in ballot (either for this election or by having joined the permanent mail-in ballot list), or you must either early vote or vote on election day. If a mail-in ballot has been requested, Colorado law imposes an elaborate process involving a replacement mail-in ballot if you have lost track of it by election day or never received it.
The Clerk and Recorder cannot under current law simply mail a ballot to every active registered voter, and also allow anyone who wants to do so to vote in a polling place on election day.
The Clerk and Recorder is appropriately concerned about the problems involved in an all-mail ballot.
The requirement of current law that voters provide their own return postage amounts to a small poll tax that can suppress marginal voters. There is also a long history of the elections division incorrectly informing voters or failing to inform voters at all regarding the proper postage for a return envelope. Voters are required to pay their own postage despite the fact that it would actually be cheaper, according to the clerk and recorder, to pre-pay all return envelope postage and get a bulk rate, rather than paying insufficient postage bills on ballots sent in with insufficient postage.
The active voter list is also defined too strictly under existing law. Someone who was discouraged from voting by long lines in 2006, hasn't voted since because they aren't interested in local races, and flicked some postcards sent at a time when no one is thinking about elections anyway into the junk mail pile, is an inactive voter who would not receive a mail-in ballot under current law. While very few responses were received to reactivation postcards, it is also true that very few were returned as undeliverable. A voter who didn't vote in 2006 and hasn't voted since is inactive and won't receive a mail-in ballot without requesting one even if that person is a valid registered voter, the person has not moved, there has been no change of address filed with the post office, and the person voted in the 2004 Presidential election. Eventually, a statewide voter registration database will be a partial solution to that problem, but the benefits will mostly accrue after the 2008 election.
But, this doesn't mean that the current super-precinct solution is a good one.
No one is seriously opposed to voting with paper ballots that are then optically scanned, as planned. The system is easily audited (which makes it hard to fake and easy to recount), makes the intent of the voter clear, and is only slightly more prone to mistakes by voters trying to cast their votes than voting machines. Colorado's long ballots make the old fashioned and less accurate alternative of hand counting ballots impracticable.
But, the City doesn't have nearly enough scanners or space or volunteers to count the massive surge of ballots expected in a Presidential election in a timely fashion, and even the Clerk and Recorder does not appear to have decided if the ballots will be counted locally in the super-precincts, or centrally. Colorado law is unclear about the procedures that apply if paper ballots are counted centrally, rather than at precincts, which leaves open the possibilities of challenges or problems related to the novel approach of counting paper ballots cast of precinct voting places centrally as one does in a mail-in ballot scenario. But, if the City can't afford to put an optical scan reader in all 175 precincts, it has no choice but to count ballots centrally, and it doesn't have the equipment to do a precinct based count with optical scanning machines now, about nine months from when early voting starts in the November election, and just a few months from the August primary election for state and local offices, in which many races will be contested in Denver this year, mostly due to term limits that have created open seats.
There are internal problems for the election division as well. The old Denver election commission was abolished and replaced with an elected clerk and recorder after the disastrous 2006 election (which fortunately probably didn't change the outcome of any races despite the problem because key races weren't close enough for the problems to make a difference). But, the November 2007 election while it didn't have large numbers of discouraged voters (in an off year school board election with no statewide ballot issues), did utterly overwhelm the ability of the election division to count the votes despite the relatively minor scale of the election, cast doubt on the ability of the election division to manage its largely volunteer staff of election judges, and did reveal continuing serious problems with information technology management in connection with the elections.
Denver's Election Division has nine vacancies in its quite small office (about 16 FTE) and is in the midst of a wholesale reorganization. It admits frankly that the office currently lacks the resources to count ballots in a timely manner at this time in a Presidential election with heavy election day voter turnout.
In sum, it is unambiguously clear that the extremely important 2008 Presidential elections in Denver will involve big changes with an untested group of people implementing them. And, this is a warning sign for a high probability of disaster.
Why does this minutiae matter?
All these little details matter because the partisan breakdown of Colorado's elections for President and U.S. Senate are largely a function of turnout in heavily Democratic Denver.
Of course, massive voting machine decertification and state voter registration problems may also screwed up many other counties in the state of varied partisan makeups. But, most small counties in the state have small numbers of experienced voters who have stable lives and whose turnout is very high. Their small scale also permits the election process to be kept simple. And, other large counties also, by and large, have less fickle voters less likely to be tripped up by minor or subtle flaws in the system.
Denver's voters, in contrast, tend to be the least experienced and most fickle in the state. Tens of thousands of Denver voters choose to vote only in Presidential elections, and are easily dissuaded from voting by relatively minor barriers to voting. Relatively few Denver voters request mail-in ballots. Marginal voters are overwhelmingly Democratic leaning. No other county in the state has such a daunting get out the vote challenge or a great need to have a smoothly running easy to use election system.
Colorado has historically been a "purple" state, so the likelihood that the top of the ticket partisan federal races for President and U.S. Senate in Colorado will be close enough for Denver turnout to make a difference is high. The U.S. Senate is closely divided on partisan lines, so Colorado's vote could easily tip the balance of that chamber of Congress. The last two Presidential elections have been decided by razor thin margins, so there is a real possibility that states like Colorado that have been swinging from red to blue across the board, could tip the balances in the Presidential election as well.
In other words, the ability of the Colorado Secretary of State, the Denver Clerk and Recorder, and the Colorado General Assembly (aided or hindered by the veto pen of the Governor) to get their act together to make the 2008 Presidential election run smoothly in Denver, could decide which part rules the critical levers of power in this country for years to come.
Alas, this time around, the seemingly simple ministerial task of running that election is anything but, and securing the smoothly implemented high turnout election which should be a foregone assumption will be anything but. We will be nail biting until the very end this year.