November 10, 2006
Regional vote centers cannot be an afterthought

Elections went smoothly in many places around the nation on Tuesday, but Denver wasn't one of them. This year Douglas County, one of the counties in which Denver is located, tried out the new regional voting center model pioneered successfully by Larimer County, Colorado.

That success didn't translate to Denver. In fact, the Denver Post called it "a total fiasco." It was so bad the mayor of Denver offered to pay the parking tickets of voters who had to wait for hours in line.

I don't think the lesson of the fiasco is that voting centers are an unworkable idea, but that they are not to be approached lightly. Apparently the main hang-up was that too few computers were available for finding a voter's registration, a necessary requirement in a system where any voter from any precinct can use the location. That's just one of the new processes necessary to make regional vote centers work.

King County is planning to use similar regional voting centers, perhaps as soon as the 2008 presidential election. While they will have less traffic than those in Denver, the problems in the Mile-High city shouldn't be ignored by Ron Sims and Co. Voting centers aren't just glorified polling places; they are an entirely new concept that must be planned and tested very carefully.

Posted by Jonathan Bechtle at November 10, 2006 12:22 PM | Email This
Comments
1. Hi Jonathan-


Good catch. Thanks for posting about this.

Since we still don't know what the plan is for forced mail voting, including the high volume optical scanners, automatic signature verification, new facilities, etc., it's hard to criticize.

Regarding the regional voting centers, I'd like to know how big they'll be. My concern is that they won't be large enough or enough to go around.

King County had nearly 32,996 provisional ballots cast. Then add the mail drop off, disabled, and misc poll site voters.

My poll site handled about 480 voters (paper ballots, touchscreen ballots, mail drop off, and touchscreens). We had 5 poll workers (including me) and we were slammed.

It took voters 8 - 12 minutes to vote with a paper ballot. Add another 1 - 2 minutes for touchscreen voters (because of the additional paperwork). Voters who didn't come prepared took much, much longer. Though I didn't time them, I'd guess provisional ballots took about 20 minutes, start to finish.

So, given these stats, historical turnout levels, and estimates, I would to know what the plan for accommodating everyone is.

I'm also curious if our new super poll sites are permanent or temporary. My guess is permanent. Our current poll sites are rented for a whooping $10, racking up a massive $5,120 bill for every election. (More or less.)

I'm having a hard time imagining the super poll sites being cheaper.


Cheers, Jason

Posted by: zappini on November 10, 2006 03:06 PM
2. Crud. I forgot to link to King Co's election results page. The 32,996 provisional ballots cast figure comes from the 2004 general election.

Posted by: zappini on November 10, 2006 03:08 PM
3. The reality is that if we are to have the sort of high tech election gear and low human error factor demanded by the MSM and the victimology lobbies, then the traditional polling place model must change - and the only practical alternative is a more limited number of regional voting stations (probably also with an early voting period, essentially changing election day at thousands of polling places to election fortnight (or so) at a few dozen polling centers).

The cadre of election judges generally consists of senior citizens volunteering for 3 or 4 days every 2 year cycle. They do a great job, all things considered. But the process is getting much more complicated and the tolerance for the inevitable errors is getting lower. The current model doesn't allow the time/experience/learning curve factors to work out to the level of error free expertise needed to operate the new election process. More regional polling centers operated for 2 or 3 weeks allows for the improved training and working time needed for the smaller pool of poll workers to implement and oversee the system with the needed lower error rates (it also gives greater protection against system failures - if there is a crash on one of 21 days, it allows for people to work a return trip into the mix and still be able to vote).

Posted by: krm on November 10, 2006 05:03 PM
4. Regional voting centers should suppress the provisional ballot turnout in King County when the county goes to all-mail voting.

Provisional ballots are typically used: (a) when a voter is inactive status, due to moving without notifying the Elections Dept of their new address, (b) when a voter registers right before the deadline and their registration isn't on the computer in time, (c) when a voter misplaces their absentee ballot, or (d) when a voter can't figure out the correct polling place.

In November 2004, there were 28,010 validated provisional ballots cast in King County, out of some 32,996 that were cast. Voters were allowed to cast provisional ballots at any of more than 500 polling places.

The new vote-by-mail plan, adopted by a 5-4 party line vote in the King County Council on June 19, 2006 in Ordinance No. 15523 (Democrats for, Republicans against), calls for a handful of regional voting centers (assuming the county can afford to operate them):

http://www.metrokc.gov/council/committees/elections/vote_by_mail.pdf

The alternative would be to allow provisional voting only at the County Elections office at the county courthouse -- something that Snohomish County and most of the 34 vote-by-mail-only counties required this year.

I would imagine that provisional ballot voting is much lower in every single one of the all-mail-voting counties this year, than it was in the last comparable midterm election in November 2002. Voters simply aren't going to be highly inclined to travel many miles to the county courthouse to cast a provisional ballot.

Even if there are several regional voting centers that can be used for provisional voting, and even if they are sufficiently staffed for the flood of provisional voting that takes place in presidential elections, the rate of provisional voting in King County should be significantly suppressed in November 2008 from what it was back in November 2004.

Had the Democrats' all-mail-voting plan been in place back in November 2004, there is no doubt that Dino Rossi would be Governor today. The provisional ballots in King County had a higher Democrat margin than King County did as a whole. If provisional voting was suppressed even by just 10% from being made more difficult, this would have removed 2,801 votes from the King County tally. Even just using Gregoire's 18% overall margin in King County, this would have been a net reduction of 504 votes for Gregoire, or nearly four times what her margin of victory was.

Obviously, the Democrats must not have thought about this all-mail-voting plan very much. If they realized how much it would suppress Democrat provisional ballot voting, they would have never proposed it, much less approved it.

Just imagine the furor that would have ensued if Ken Blackwell had restricted the places provisional ballots could be cast in Ohio, or if Katherine Harris had done the same thing in Florida. They would rightfully be accused of trying to suppress Democrat votes.

But for the King County Democrats to score an own goal, and get rid of perhaps 10,000 or more heavily Democrat provisional ballots -- PRICELESS!

Posted by: Richard Pope on November 10, 2006 05:53 PM
5. Hi krm,

What kind of errors occur at poll sites? Is that error rate more or less than with mail balloting?

The Caltech/MIT Voter Technology Project did a study back in 2001. They concluded that the least error rate comes from using paper ballots at neighborhood poll sites cast on voter correctable precinct-based optical scanners.

Maybe you have better or more recent data.

Cheers, Jason

Posted by: zappini on November 10, 2006 06:51 PM
6. Jason -
A lot of errors are possible in the polling place that don't show up in studies of error rate by voting format.

For example, not properly setting up or booting up the equipment. That can lead to not opening on time, which leads to long lines (in the before work 'spike' of turnout) and people deciding not to wait out the lines (and perhaos not to come back at teh end of the day in the event of a court ordered additional time period tacked onto the end). By way of example, only 136 of 233 precincts in Kane County, Illinois (suburban Chicago) opened on time on November 7, 2006. Most were no more than a half hour late, but a few were more than 2 hours late getting going. Very little defective equipment factored into that county's problems.

It is not really possible to provide an "error rate" in the traditional academic sense with regard to the people who did not vote at all because of polling place opening problems.

I've done some election judge training. They are almost all wonderful people. But some of them still have their VCRs flashing "12:00" and are not, operating under pressure of a few hundred impatient voters, going to really master three, four or five pieces of what is essentially (if not literally) computer gear.

Posted by: krm on November 11, 2006 01:10 PM
7. Denver is a city/county all its own. It is not part of Douglas county. Douglas County is a rapidly developing county between Denver and Colorado Springs.

That said, regional voting centers were a disaster in Denver. And the people staffing the election commission knew it was coming.

Douglas County didn't do so well either. Waits to vote were very long, people stayed until midnight in some cases even thought the polls closed at 7.

Posted by: LG on November 12, 2006 06:08 PM
8. LG: Thanks for the correction. The media reported problems in both places, and I misunderstood the connection.

Posted by: Jonathan Bechtle on November 13, 2006 02:57 PM
Post a comment
Name:


Email Address:


URL:


Comments:


Remember info?