undefined | Better HN

0 pointsCaliforniaKarl8y ago0 comments

I was a polling place Inspector in Orange County, CA for around six years.

In the lead up to an election I’d work, first up was ~2 hours of training. We were the ones in charge of the polling place, so we were the ones who needed to know all the processes and procedures.

The weekend before the election, I would pick up the supplies. Probably around 40-50 pounds of stuff. You aren’t allowed to leave In in your car: Once you sign the chain of custody for it, the next time you get out of your car is when you’re parked at home, ready to unload.

In the weeks before the election, I also would be getting in touch with whomever is in charge of the polling place site. You think it’s fun dealing with an HOA as a resident? Well, try dealing with them as an outsider.

So, leading up to the election (that is, prior to E-1), I would’ve logged about 7 hours, most of which was during a business day (training could be daytime or evening, as there were always multiple sessions available, and supplies pickup was always on a weekend).

Oh, and I was also responsible for managing the other people would be at the polling place.

On E-1, if I got prior OK from the property, I would set up some time during the evening. Sometimes one or two of my clerks would be available to help, but not always. Anything non-sensitive and nonessential was set up (so, no breaking seals yet!).

On E-Day I would be waking up around 4:30. Polls open precisely at 7, so I had to be at the polling place by 6. I would take attendance, administer the oath, go through the same oath myself, and then do all the seal verification and equipment setup. Then, the flood began.

If I had a full board, I could give everyone (and myself) a lunch and a break. If we were short, then we would do what we could. Luckily, we all got some sort of break!

There would be quiet times during the day, but it would often be nonstop during the morning (there was always a line at opening) and after 4.

Polls close at 8. We then had to do full packing, space cleanup, and all of the accounting. It was a win for us if we left by 9.

I then had to go to drop off everything, with a clerk driving behind me (to make sure I went directly to the drop off point). If anything was missing, I was on the hook for it. It was a win for me if I was out by 10.

Finally, on E+1, I’d have to return the facility key. Then I was truely done.

It was a drawn-out process, but almost everything we did had a very good reason. Yes, the hours were long, and the work was often not fun, but I took a _great_ deal of pride in it.

0 comments

3 comments · 1 top-level

javajosh8y ago· 2 in thread

Thank you for your service, Karl. Perhaps you might share some more light on your experience "administering the vote" as a polling place inspector. As a technologist, what do you see as the major error modes of the American system of polling? Do you think fraud is wide-spread? Is it even possible to measure fraud (since, by definition, if you detect it you eliminate it)? Are there other methods of polling that might work better? It seems like a really hard problem, and I've wanted to ask someone "on the inside" for a while now, and this seems like a good opportunity. Thanks in advance.

CaliforniaKarlOP8y ago

Hello!

To be clear, I wasn't an employee of the Orange County Registrar of Voters, so I wouldn't count myself as being "on the inside". Neal Kelley, who was Registrar of Voters back when I volunteered, is still there today, and was doing alot of stuff with YouTube at the time. Check it out: https://www.youtube.com/user/ocrov/videos

There is also a _ton_ of poll worker training info online, which you can read here: https://www.ocvote.com/volunteer/poll-worker-resource-center... The dates say 2016, because there won't be many updates in off years.

I first volunteered because Orange County were using the Hart InterCivic eSlate voting system. You can read about them here: https://www.verifiedvoting.org/resources/voting-equipment/ha...

These systems didn't run Windows. They ran some sort of minimal OS that had a sub 10-second boot time. Power and data went through the same cable, with a DB-9 connector at each end. All case seams and unused ports had security seals on them, which were checked at pickup, at dropoff, and during the day (we had chain of custody paperwork that travelled with the hardware). All the voting machines had printers. The eSlates were in stand-up cases with all the necessary hardware (feet for standing up, privacy shroud, etc.), and which also kept all the ports covered up. Since power to the eSlates came through the data cable, anyone messing with a cable would take out their eSlate, and all the ones after it.

(NOTE: You may notice many setups having a power cable for each eSlate. That's not a power cable for the eSlate, it's a power cable for the printer. The printers are thermal printers, and draw enough power that they can't use the daisy-chain power source.)

All of our eSlates had printers. Ballots lived in each individual eSlate's memory, in the JBC's (the controller's) memory, and in paper form (the printer was sealed). It could run on battery for a limited time, though we couldn't use that, because the printer wouldn't work. We had a supply of paper ballots if all else failed. Emergency support (new equipment, more paper ballots, etc.) was available typically within an hour.

I ended up really liking the voting system, because it was simple on the face of it, but it was also clear to me that it wasn't using a Windows OS. I wish it was used more.

> As a technologist, what do you see as the major error modes of the American system of polling?

You're not going to get the answer you want. I know alot about the executing of the voting process for Orange County, California. That _might_ translate to other counties in California (as election policy is set most at the state level), but I'm not going to speak to other locales. To do so, I would need to study the county's training material, and see polling place-level records of a previous election.

But, _speaking specifically to Orange County, California_, I noticed two issues:

1. Change of Address often did not make it to the voter rolls. If you moved within a county, it was more likely to process OK, followed by moving between counties, but moving between states often didn't remove you from your former home's rolls.

I think this was a trust or coordination issue: Since policy is mostly set at the state level, it's easier to coordinate information transfer between counties in a state. There isn't much in the way of a Federal ID number, except for the SSN, and I doubt your Registrar would want anything to do with that number.

2. Your full ballot is most often only available at the polling places within your precinct. IMO, that is because of (a) the voting systems are not using full-featured OSes, so they lack the capacity to store that many different types + languages of ballots; plus (b) each polling place only has their own rolls, so your info is only listed at your polling place.

Note how both issues seem to be forms of information synchronization problems: Transferring data between regions (which doesn't have to happen quickly), or synchronizing voter sign-in data between polling places in real time.

Personally, in my opinion, I would rather _not_ have real-time sign-in data sync between polling places, because that adds more layers of technology that could go wrong.

Also, we had a good method of handling provisionals: If anything would go wrong procedural-wise, we would record information about you, and us (the polling place), and the situation. You would then sign paperwork saying that the information is correct, and you haven't already voted. You would then vote, with your ballet uniquely IDed. All the info, and the unique ID of the ballot, would go into a sealed envelope, which was treated like a voted ballot. Your vote would not count in the immediate numbers (the numbers released that night), but if everything checked out, it _would_ be counted in the final, certified numbers.

According to https://www.ocvote.com/results/left-to-count/provisionals-le..., OC had ~130k provisional ballots processed.

>Do you think fraud is wide-spread?

No, I do not think fraud is wide-spread (_in Orange County, California_), precisely because of the provisional process. For example, if someone voted by mail, the rolls reflect that. If someone then comes in and tries to vote using that name, then they have two choices:

1. Hand over your un-mailed, empty return envelope, and vote at the polls. We then void your envelope, and keep it in a special container, separate from everything else.

2. Vote provisionally.

(If you have a sealed, ready-to-mail return envelope, you could also drop it off to us directly, to save the stamp. We treated those the same as voted ballots.)

If anything appeared unusual, we would record all the information, and then the voter would vote provisionally.

It's worth noting that the provisional voting process added 5-10 minutes to the voting process, so perpetrating fraud would need a significant investment in manpower, with no guarantee that your provisional ballot would make it past the people at the ROV who would examine it.

The only real way I could think of to perpetrate voter fraud is to memorize a significant number of non-vote-by-mail voters' full names and addresses, and then claim to be them at the polling place. But, again, you need alot of manpower, because it's the same Clerks for the entire day, so if you show up multiple times, people will notice. Making vote-by-mail easy to do really helps stop this, because of what I described above.

> Is it even possible to measure fraud (since, by definition, if you detect it you eliminate it)?

Yes, you can. You take the rolls after the election, scan in the signatures, and compare those signatures to each voter's original registration. If a _specialist_ can confirm a signature mismatch, then you have the _possibility_ of one instance of voter fraud, which you (by definition) _cannot_ correct, because all the non-provisional ballots are anonymous.

This is another plus point for vote-by-mail, as you can catch these cases before removing the ballot from the envelope.

> Are there other methods of polling that might work better?

This question is too general for me to answer. For example, the system whereby states set election policy is something that comes from the country's history as a tight alliance of individual States, so IMO it's not really worth arguing that.

I think there could be two major improvements:

1. Make it _much_ easier for change of address info to get to Registrars of Voters. For example, when you terminate a rent/lease, or submit a Change of Address form to the USPS or DMV/BMV, that should trigger an automatic notification to the ROV for the county you are leaving. You would then take responsibility for registering at your new address.

The concern there is privacy, because it makes voter information even more valuable. That brings me to suggestion #2…

2. Have all voter communications go through the Registrar.

Instead of offering voter rolls for sale, the Registrar should serve as a conduit for mailing/calling.

I mean, companies like Facebook let you target ads towards specific groups of people. The Registrar knows who voted (though not who they voted for), and they know people's addresses & party affiliations. They also have the infrastructure in place for doing large mailings (since they mail out ballots), so if you're a candidate and you want to mail stuff to certain people, give the materials to the Registrar and have them do it.

makomk8y ago

For context, there are heavy legal restrictions on why, how, and when moving between states in the US can remove people from the voter roll in their old state. Unless the voter tells their old state to remove them, the state has to send them a notice and wait until after the second Federal election - if they vote in any election before then, the state can't remove them. (This is probably why Gregg Phillips, the guy who Trump was promoting as a voter fraud expert for some reason, was registered in multiple states.) Also, the only legally safe trigger for starting the removal process is the US Postal Office's change-of-address process, and it wouldn't surprise me if the ACLU challenged even that.

Not only that, removing people from the voter roll for this reason almost always results in headlines implying or outright stating that this is voter suppression. There's even a popular conspiracy theory about the 2016 US presidential election (started by journalist Greg Palast) that relies on the assumption that almost every voter marked for removal from the rolls in the key swing states of Pennsylvania, Michigan, and Wisconsin in this process was a Democrat voter whose vote was suppressed.

j / k navigate · click thread line to collapse