It’s pardonable to be defeated, but never surprised.
— Frederick the Great
Over the past week, there has been plenty of harsh criticism of the Romney campaign’s get-out-the-vote efforts and volunteer system. While the campaign has pushed back on these claims, my experience inside a swing-state legal war room last Tuesday leads me to say that the criticisms of the campaign’s GOTV in general, and of its use of the Orca system in particular, ring true.
The main GOTV effort was supposed to operate like this: Poll watchers listen to the names of everyone coming in to vote and notate them using Orca. This means that other campaign workers charged with calling folks and encouraging them to vote don’t waste time on those who already have. It should be an easy way for campaigns to make efficient use of limited resources.
I know a few things about Election Day efforts and election law. I was an election judge for five years, and I have also been on roving legal teams. Based on that experience and on what I saw on Election Day, here is my judgment as to how and why the Romney campaign’s efforts failed, and what future efforts should do differently.
Prior to Election Day, the training for Romney poll watchers consisted of large conference calls led by staffers at the Boston headquarters, each with an enormous number of volunteers calling in. Such top-down fora are the completely wrong way to do this kind of training, because they discourage volunteers from participating and asking questions. This left questions unanswered, which caused big problems on November 6.
I witnessed particular confusion about the Orca system. It was basically a Web page that the digital team called an “app,” leading volunteers to think they needed to download something only to have their efforts hindered on Election Day morning.
In the future, it needs to be clear exactly what volunteers will be looking at on Election Day. Each poll watcher needs to go into the precinct knowing exactly what to expect.
Imagine you assigned an employee to give a presentation to a key client. Would you keep the content of the presentation, the makeup of the audience, and the technical aspects of the projector and computer he was to use a secret from him until the day before the presentation? Of course not.
Yet that is basically what the Romney campaign did to its poll watchers. It asked them to show up on Election Day to run a new system that they were not allowed access until November 6. Even if the system had worked perfectly, this would have been a serious oversight, but in many cases the Orca app just didn’t work. In our legal war room, we were flooded with pleas for help as Orca malfunctioned. Many users’ log-in PINs did not work, and the process to reset the log-ins failed. To paraphrase one president’s take on poor crisis response, one would not run a birthday party like this.
For any new system that has to work perfectly on a particular day, basic testing beforehand is critical. In this case, the system should have been tested in states that had late primary elections. In addition, any system must go “live” long before the day it is needed; in this case, volunteers should have been able to test their PIN, try loging in, and browse through the system to get familiar with it. Any person who has been in business even briefly knows that launches have a “soft” and a “grand” opening. Working out these kinks early next time will avoid another meltdown on Election Day.
Further, it seems that in the months leading up to November 6, the legal-team effort was walled off from much of the poll-watching effort. These efforts should be integrated. In my past work on roving legal teams, I was given ahead of time the names and mobile numbers of the poll watchers in each of the six precincts I was covering. A volunteer should not just be sent out by himself to a polling place with no sense of support and no idea whom he could turn to for help. Understandably, without any support line, many poll watchers went home when their efforts to address the Orca meltdown were unsuccessful. There were many important precincts in swing states that ended up without a Romney poll watcher because of this.
I suspect that if they could have texted that nice campaign aide or attorney who had called them on Sunday, they would have done so rather than giving up. The failure to connect volunteers is especially exasperating when you consider that the entire point of GOTV is to identify and personally reach out to voters, on the premise that personal contact is most effective. Why would the same not hold for these wonderful, dedicated volunteers working the polls?
When packing for a trip, the most important thing to remember is to bring your photo ID; everything else can be replaced at your destination, but your ID cannot. On Election Day, a poll watcher’s ID is the certificate that enables him to get into the polling place. No certificate, no poll watching. Every state was different, but if there was a weakest link in the whole Romney GOTV system, it was the certificates.
On Election Day our war room was getting calls, to our confusion and surprise, from volunteers who either had not brought their certificate or had not been given one at all. The certificate should be the most basic point in poll-watcher training; it was clearly not a top priority for the Romney team.
One failure we saw firsthand in the war room was the system’s showing individuals as having voted absentee when in fact they had not. And I don’t mean occasional instances, but entire lists of people. Part of the poll watcher’s job is to alert the legal team if he suspects any kind of fraud, and having an absentee voter walk into the polls to vote again is a classic example of voter fraud. But if the lists the poll watchers are given aren’t accurate, we will continue to have poll watchers raising questions about legitimate voters. Fortunately, election officials, as they should have, used their authoritative registrar’s list so there were no real confrontations. But this was still an embarrassment and a waste of time. I hope this problem was isolated to the swing state I volunteered in, but it would surprise me if we were the only ones.
I was very surprised to learn — the night before the election — that Orca would be run from headquarters in Boston. When the Orca system was down, our legal war room was flooded with pleas for help because the hotline in Boston wasn’t working. We even had calls from volunteers in other states asking us for help getting through to Boston. We were combat ineffective.
Having everything controlled by a national headquarters is just asking for trouble. Election laws are different in each state, and the training and Election Day operations should be organized at the state level. Furthermore, if one swing-state office’s power, Internet, whatever, fails, the damage won’t affect other states’ operations. Having a national number as a last resort is fine, but relying on a single central node that could fail is lethal. If there had been a complete network failure the first day that Amazon.com was in business, they would have fixed it and been back in business the next day. But you can’t postpone Election Day, and there are no silver medals for finishing second.
From what I have heard, the angriest critics are Romney people who have privately expressed disbelief that others on their team would let the system be designed in such an unreliable and vulnerable way. It is still unclear to me why the GOTV wasn’t done by the RNC — it’s a long-term effort and a natural role they have performed well in the past.
The Election Day failure was as if a team had spent a year building a car to beat the land-speed record, but when the green light flashed on the Bonneville Salt Flats, the car didn’t start because there was no gas in the tank. It is up to us to make sure it does not happen again.
— Michael James Barton is a director at ARTIS Research, and has served in a variety of leadership roles on Capitol Hill, the White House, and the Pentagon.