bruce, computers may be more reliable--but humans are more accurate. A computer will count a statutorially valid vote wrong, the same way every time.

No matter which way this goes, officials are not taking a second look because of "sabotage" - or any other charge of criminal activity or fraud. And if big money narrowly buys their way onto the ballot, I say bring on the debate - a victory at the polls will be sweeter anyway.

One thing everyone has to remember is that the counties are now using a new state-wide system for voter registration. Mult Co did some testing of it. But that doesn't mean everything is going to work perfectly. You'll find errors along the way and get them corrected.

They'll fix whatever is wrong and then look at the petitions again to see if there was any affect.

You'd better believe that the county is being as careful as possible with this, as they know that they're likely to be sued if they don't. No one spends that much money on a petition effort without having a plan for legal action, if needed.

what's the problem with sampling? I don't get it. It's an entirely valid way to approach the counting process scientifically.

Ramon, once again I urge you to provide information on how sampling "manipulates" the outcome. So far you've failed to do so. Simply repeating "Sajo v Bradbury" doesn't cut it.

LT--the answer would be yes and no. If the machine reader uses optical scan technology (bubbles you fill in, essentially), then it WOULD pick up a double vote like that, IF the 'write-in' category had its bubble filled in. Unfortunately, people who do write-ins have a tendency not to fill in that bubble.

Here's an example of a valid vote that machines will reject, "reliably," every time: voter selects Candidate A, then crosses it out and writes NO next to it, filling in the bubble next to Candidate B and writing THIS ONE next to that. In most jurisdictions, this would be a valid vote for Candidate B. However, the machine would almost surely reject it as an overvote.

Fearful isn't the word. "wondering why the time and expense are necessary" would be more apt. The only data I've seen indicate the repeal wouldn't pass, so I'm not sure why one would fear it.

Perhaps the fear you're referring to is the fear that a select group of large business interests are acting out of their OWN fears, and will spend whatever is necessary to preserve their special access.

Does anyone else find it ironic that those who most fervently support voter-owned elections are the most fearful of letting the voters decide if they want it?

Good one, Bill. I support VOE, I guess that's why Im so "fearful" that I wrote above:

And if big money narrowly buys their way onto the ballot, I say bring on the debate - a victory at the polls will be sweeter anyway.

No Bill, it's not a bit ironic and I'm pretty sure that you understand why.

The soundbite that will be used against VOE is simple and deceptive: "Are we gonna let those sleazy politicians spen our money to get their fat cat butts reelcted?" (You know the argument, some of your guys are actually brain dead enough to try to float it on the informed citizens at Blue Oregon).

The argument for VOE is that it is wa-a-a-a-a-ay cheaper in the long run to get the Big Money Kids out of the election funding bidness, because they almost always expect a quid pro quo for their donations, and whatever breaks they get have to be paid for by the rest of the taxpayers. Then you have to explain how this all works.

Any group that is facing a simple, great sounding, but fundamentally dishonest argument, would prefer not to have to deal with it.

Don't worry, though, I'm sure that when the first Gard and Gerber cowpie hits the rock, there will be hard eyed border collies armed with the facts, doing their best to drive off the coyotes, and protect the herd.

The expensive nature of checking every signature on the petitions has nothing to do with the databases-- it has to do with the number of hours you have to pay people to check those signatures.

It would mean either overtime for the people already working there, or the need to hire and train additional people.

As far as the questions about ballot with things written on them...

Ballots with extra items written on them are checked by hand. If it can be seen how the voter intended to be voted, it's counted (this is done by a group of people from different political backgrounds). If not, it's not counted.

When working at the elections office, we got a lot of calls from people who had messed up on their ballot. We just explained that they would need to make it obvious how they meant to vote, such as writing "no" next to the one they incorrectly voted on and "yes" next to the one they meant to vote for. Or crossing out the bad vote, and bubbling in and circling the one they'd meant to vote for. We explained how they would be checked by hand. Some were fine with that, others would go ahead and come in for a new ballot.

While it may mean extra time has to be spent on their ballot, they don't have to wait in line to get a new ballot (which can take a while, since they have to verify no ballot has been returned, print up a label to go on your envelope, and pull the appropriate ballot).

If you want to know more about how the machines count different kinds of votes, just head on down to your county elections office. They go through the process of explaining this to people quite often.

There are very few employees between elections. The vast majority are temporary employees and hired only for election season.

The others have a variety of things they work on.

A few are trained in passports and also run the passport window there at the Multnomah County elections office.

Some work on the local/county race C&Es, which come in throughout the year.

There are voter registration cards that come in all year, and they input them.

Checking the signatures takes quite a bit of time. It's not like checking signatures on ballots.

On ballots, the bar code has already been scanned and they're in batches. You bring up a batch and every ballot in your box is there, in order.

With petition sheets, you have to figure out the voter's info (which if you've ever dealt with handwritten sign in sheets, you know how hard that can be), run a search, find the right voter, etc. When you have a large number of signatures from people who aren't voters (or haven't updated their info after moving, getting married, etc.), it takes a lot longer.

Also, I should note that the election doesn't take up only 2 months a year.

Actually, the busy part of each election starts more than a month before an election and can go weeks after the election.

Ballots will be going out April 28 for the May 16 election. The voter registration deadline is the day before-- April 27.

The weeks prior to this are spent inputting the large number of voter registration cards that come in prior to an election. While the number of registered voters may not change, you get a huge number of cards from people who need to change their information.

There's also all the planning that goes along with an election.

Not to mention all of the work that surrounds candidates-- filing for office, C&Es, verifying signatures for those who get on the ballot that way, etc.

For the primary election there's also the need to check the information on everyone who files to run as a precinct committee person, which is more than 1,000 people over the two major parties.

There's a lot more to the elections office than you may think.

There's a lot more than 4 months of f/t work. There's about 3.5-4 months of work during the 2 election cycles in which temporary workers are brought in to help with the heavy load.

That doesn't include the time when the temp workers are no longer needed because the work load drops enough for the permanent staff to handle (or at least handle with only a handful of temp workers).

Then there's all the other work that they do, some of which I mentioned above in the previous posting. That's definitely not all they do, as I know several people there also handled other things (like they do with passports). I can't remember for sure what those are, so I won't list them here. But you can call the head of elections and quickly find out.

Outside of the management and admin positions, there's only about 8 or so employees there the rest of the time. And every single one of them has duties that they do which fill the remainder of the year (such as handling passports, C&Es, etc.).

Adding information to the signature sheets would be helpful. DOB would be good, as that makes records easy to find. Many people don't know their voter reg number, and they'd have to drag out their DL to get that number.

You'd still have the problem of having to read their handwriting, though, which can seriously slow things down.

I haven't heard of the rules allowing for a group to be allowed to pay for all signatures to be counted. If they don't, this could be because it gives organizations with loads of money a significant advantage over small groups that entirely rely on volunteers for their signatures. The goal is to put everyone on an even playing field.

The bar-code reader would end up having to be something the gathering organizations provided-- I don't see the county or state governments buying and handing these out. Even then, you'd need something that would allow you to personally sign, sort of like what the UPS guys use. I think seeing the person sign the petition, digital or not, is a very important part of the process.

And of course this once again puts the volunteer powered groups, who often can't afford to spend hundreds of thousands to hire a company that do signature gathering, at a disadvantage.

I never mentioned the words "bond" or "recount." Those words were use by others.

I'd imagine those tables like the UPS has cost about $100 each. Of course that doesn't take into consideration the hardware and software that would also be needed-- that's what can get expensive. And they work by scanning a barcode on the package, which then brings up a place for you to sign your name. The needed tablets would have to do more.

With those tablets, you'd run into the problem with people who have moved and have the wrong information in the system (people are not good about updating their address). If they've moved, and not updated their information, they cannot sign a petition. They have to update their information first (which they can do right there with a voter reg form). Just swiping their ID and then having them sign would lead to a good number of signatures with addresses not matching the voter reg database. And of course that would mean a lot of wasted effort by those collecting signatures.

A better system would be similar to the way some surveys are done now-- with a laptop. The information could be quickly typed in, and there could be a digital pad where people could sign their names.

But even if purchased and then rented out, they'd still be out of reach for the smaller organizations, unless you made it all very cheap. But then you'd run the risk of losing money if items were lost or damaged.

If the county is going to have to do overtime (elections offices are run by the county), then the petition gatherers should have to pay for it. Right now the taxpayers already foot a large amount of money on these efforts-- checking the ballot language, approving the petitions, checking the sample, printing the voters guide (which does not pay for itself), running the election, etc. I don't think people are going to be too happy about how much more expensive this could get unless those gathering the signatures pitch in-- and they don't mean the amount of money spent hiring people to gather signatures, lawyers to write their language, etc.

I've been involved in election administration and reform work for about two years.

The Oregon elections bureau run by John Lindback, and under him the Multnomah County operation run by John Kauffman, are nationally known for their quality and expertise. We are national leaders in many aspects of election administration.

If you think his staff is 'fat' and wasting time, then head on down to 11th and Morrison and take a look. Seriously. John K. has always been willing to let anyone watch how election administration operates. it's fascinating to watch the signature verification process in action.

So while someonen may object in principle to using sampling as a way to verify petitions, you're off base if you think the Elections Office is incompetent, inefficient, or controlled by party hacks.

Having them change the information there in the system isn't the only thing they need to do-- they'd also need to fill out a voter reg form.

You need to have a form filled out with a signature. The form can be completely filled out via computer (such as the fill-in PDF), printed, and then signed. How it's filled out (by hand, someone else doing the writing, printer generated, etc.) doesn't matter-- it just matters that the person verifies the information is correct and then signs it.

The form is important, as they are scanned into the computer and then stored. Sometimes those original copies are needed for various legal reasons.

Of course, that still means you're likely to end up with a number of "invalid" signatures from people who have moved, updated their DL, but not their voter registration. After handling hundreds of these per day during the 2004 election, I can tell you the number of such people is astounding. People think that when they change their address with the postal service or at the DMV that their registration is changed as well. The only way to really solve this problem is to educate people or change the system. There's not much we can do about the postal service change of address, but maybe with the DL change of address.

As far as the staff goes...

Their load isn't all that lighter-- it just means they're not regularly working over-time. In the off season the amount of work they do have can be handled by a smaller staff and temp workers aren't needed. But there is more than enough work to fill a work week for all the employees.

To check all signatures, they'd need to bring in temp workers which would cost the county money.

Bruce, I still haven't heard you explain what your beef is with sampled evaluation. What part of the science of statistical sampling don't you agree with? The research is voluminous and well-tested: a truly random sample of a known universe yields highly accurate results. The known universe--signatures--is a given. The only other consideration is choice of sample, and I've not heard anyone offer evidence that that part of the process is compromised.

As for exit polls--funny you would mention it that way. The fact is that exit polls are VERY accurate relative to the ballot counts. That's why they are the tool used by elections monitors to determine whether elections in other countries have been run fairly. And it's why the exit polls for President '04 continue to receive scrutiny--something is not right with them.

Please--some of us went through this exact same nonsense when Dino Rossi challenged the Washington goobernor's results. Jenni has done an outstanding job explaining to you just what goes into a process like this; I suggest you read what she writes carefully. It matches up squarely with everything else I have learned about how good municipalities conduct their elections.

Finally, as to timing: I believe Multno is compelled to respond to the submission within 15 days, because there is more to the process than just verifying the signatures. Ballot titles need to be created and ajudicated if necessary, printed, etc.

Bruce--600 signatures is not razor thin--it's 2% less than they needed. Electoral recounts don't typically happen unless the margin is less than 1%, and often is .5% or lower. And it's certainly larger than the error margin, which would be tiny indeed given a large 20% sample.

Your rant about electoral votes and Al Gore is entirely irrelevant; I never said anything about popular votes or even the 2000 election. It is not a conspiracy theory to say that exit polls are the standard by which fair elections are judged worldwide, and that they are highly accurate in determining election results--partially because they self-adjust as more factual information comes in.

As I stated in response to your first paragraph, the margin of "defeat" is much larger than the margin of error from a 20% sample.

LT--my reference was to national exit polls for the executive. I never even approached saying they were infallible; I said they were highly accurate determinants of actual results.

Using polls as predictors is never entirely wise; especially the further from the polled event you are if it's a horserace poll. But they are valuable as both point-in-time and trend indicators.

I stopped reading WBA2's comment around the Pinochet comparison, but a strawman keeps coming up: "VOE reformers are just afraid to face this repeal."

Earlier in the post, I wrote:

From Bill: Does anyone else find it ironic that those who most fervently support voter-owned elections are the most fearful of letting the voters decide if they want it? Good one, Bill. I support VOE, I guess that's why Im so "fearful" that I wrote above: And if big money narrowly buys their way onto the ballot, I say bring on the debate - a victory at the polls will be sweeter anyway.

Hey if PGE, Qwest and the FTF committee got the repeal on, great. Game on. If anyone is afraid here, it's the big money folks who are afraid to give this system a shot. Methinks they fear it will work.

Bruce, again you misapprehend. It's incorrect to say the margin is based on technological and procedural errors. There are errors in the voter db; as yet it is unknown whether those errors affected the outcome. For example, let's say the voter db was missing 1000 valid registrations. For it to have any impact, some of those missing registrations have to ALSO be people who signed the petition.

I did not call this process a "recount," I was referring to electoral recount standards, in which a recount is generally not allowed unless the result margin is 1% or lower. It makes no sense to compare the qualification deficit (624) to the total number of signatures submitted; FtF could have submitted 40 million instead of 40 thousand, and the number needed would not have changed. To get the margin of defeat, you would compare the valid signatures to the signature threshhold. The margin of defeat is thus 2.34%--well above the threshhold considered "close" by common electoral standards.

The sampling protocol is unaffected by the voter db errors, as far as it appears. Choosing a set of signatures to review has nothing to do with which ones are in the voter db--if they appear, they are randomly distributed throughout the full set of signatures.

You have manufactured a "cloud of impropriety" all by yourself. No evidence has been offered--by you or anyone--to suggest impropriety by Multnomah County or the Sec of State in implementing the voter db. So quit casting aspersions unless you have something besides speculation to work from.

I agree sampling is subject to a calculated margin of error. The two factors required are the size of the universe (in this case 40K), and the size of the sample (8K). Thus the sample proportion is 20%. Working loosely from survey error margins, where a mere 1,000 responses stands in for the 150 million or so registered voters in the US and yields about a 3% MoE with .05 confidence, it should be clear that the MoE on a 20% sample is much, much less than 3%. The errors of validity have NOTHING to do with the error margin.

It is entirely possible that the fix to the voter db will result in a WORSE outcome for FtF. How? If any of the signatures originally "missing" from the db turn up as valid voters--but are themselves duplicated in the sample--then they will not only NOT count as valid signatures, but will penalize the total as duplicates.

It is the "complicated formula" that allows a sampled set of signatures to stand in for a review of the entire universe of signatures, so I don't get your complaint there, either.

Well, it's moot now...

The City of Portland announced today that the campaign to repeal the City's public campaign finance program has submitted 25,855 valid signatures, 836 less than the 26,691 required to qualify for the ballot.

The technological error mistakenly excluded 49 registered voters from the valid signature count, while the procedural error mistakenly included over 180 petition sheets that contained circulator errors. The technological issue in the Centralized Voter Registration System has been resolved and city and county procedures have been reviewed to avoid future errors of this nature.