Local election administrators across the country say the hardware they're using is outdated and in need of either repairs or outright replacement. As Governing reported in the July issue, public officials are delaying the necessary updates due to cost, regulatory barriers and limited options on the marketplace.

But Los Angeles County is bucking that trend. Dean Logan, the registrar-recorder/county clerk, is overseeing a process to design a new voting system unlike anything currently being used around the country. Governing spoke by phone with Logan on May 20 to discuss the Los Angeles County project.

The conversation transcript has been edited for clarity and length.

Would you start by explaining a little about the existing elections model that you're trying to get away from?

The traditional model of voting systems procurement is that jurisdictions contract with a vendor for a system that has been designed, built and certified by that vendor. So there is a profit relationship. In many cases, the equipment remains owned by the vendor and it's serviced by the vendor, with oversight by the jurisdiction.

Los Angeles County is somewhat unique in that we have a very old voting system that was developed by L.A. County government back in the late 1960s with punch-card voting. We have different contracts for the components of our voting system, but we're not tied to a single relationship to one commercial vendor operating and supporting the whole voting system.

We see value in that. There isn't a voting system that meets our needs, so that takes us out of the market in the first place. But we also believe that it's important that the voting system be publically owned and operated and that it has transparency and security provisions to ensure that voters have confidence that their vote is being cast as intended and counted as intended.

Why are you taken out of the market right off the bat?

First of all, we are the largest jurisdiction in the country, so we have 4.8 million registered voters. We have 5,000 polling places and we employ over 25,000 poll workers on election day. To scale the distribution of voting equipment over a large geographic area, to numerous locations, and to get those ballots back to a central location and have them counted and reported in a timely manner -- the current systems that have been on the market just don't have that ability. Add to that that we have to provide our voting materials in 11 different languages other than English under the Voting Rights Act. That's another requirement or design feature that none of current voting systems contemplated.

We also have a very diverse electorate and we are economically diverse. So we serve areas that are very affluent and conditioned to options with technology; we also serve areas that are dependent on public transportation. We have a homeless population that needs to be served in order to vote. It's just really a unique jurisdiction in terms of the combination of all of those elements.

Ok. Scalability and multiple languages are two features that you would like in a new system. Are there other features you would like to add as well? What about security features?

We've had by and large the same voting system for more than 40 years here. We're going to make a sizeable public investment in a new voting system and we want to be sure that is a modernized voting system, not just a rebuild of the previous model.

When you talk about security, we want to leverage off-the-shelf hardware. We want open-source software for the interface. We want to separate the process of marking the ballot from the process of counting the ballot. With the existing voting systems, that's an all-in-one system.

We want to build a ballot-marking process that has flexibility and is adaptable to the electorate we serve, for those voters who vote by mail, for those voters who might want to go to a vote center, or vote early or at neighborhood polling places. We want to give them a ballot to mark that is both intuitive and accessible. But then we want that to produce a uniform paper-based, human-readable ballot that is tabulated on an entirely different system that has no physical relationship to the device where the ballot was marked. That's a security feature that doesn't exist today.

Both you and Dana DeBeauvoir in Travis County have mentioned creating some kind of open-source election software. What are the main differences between what the two jurisdictions are doing?

We've worked closely with Dana and the project in Travis County. It's another model that I think will be useful in moving the nation toward the more modernized approach to voting systems. We're both still in development, so things are still undefined. I think the main distinctions between the L.A. County project and the Travis County project is that in L.A. County we started by designing around the voter experience rather than starting with designing a technological solution. We wanted to get the voter experience right and then to have the technology respond to what ends up being defined as the ideal voting experience. In Travis County, I think they started with a technology team that put together a technology solution. I think they put together something that will still be more a one-entity system from marking the ballot to tabulating the votes. It will be an all-in-one system. We're looking at separating those things. I think what's common to the two projects is the desire for transparency, looking at open-source code and looking at off-the-shelf hardware components.

When you say "off-the-shelf hardware components," what do you mean?

So, for instance, if the touch-screen interface is a tablet-based process and there's a commercially available tablet that meets those specifications, rather than have a company build customized tablets that are just for the voting system that will age out and have to be replaced over time, we could leverage the use of existing tablet components, printer components, all of that, and we would then load them with secure software interface and we would some disable features -- they would still require some customization -- but we don't need somebody to go out and develop a tablet or a touch screen. Those are components that exist on today's market and in fact are constantly being improved upon. We want to be sure that as those hardware components continue to advance and get better, that we have the ability to upgrade and integrate them into our voting system, rather than having to start over and build an entire new voting system every time there's new technology available.

Will you still make use of private contractors?

Our project does contemplate private contractor engagement. What we're trying to do is to develop the system and specifications for the system, separate from the manufacturing. So, instead of a vendor that will build the system, designing it around its business model and its ability to make a profit on it, we want to design it. We get the specifications and then we put it out to bid for a competitive process to determine who wants to build it, but according to the specifications that are already adopted.

This story was originally published by Governing