Kansas has focused on the wrong issue when it comes to election security. That is the message of one Republican candidate for secretary of state.

Dennis Taylor, Topeka, has served and worked in various forms of government, including the Shawnee County Commission and the administration of Gov. Mike Hayden. An attorney, he also has worked in the private sector in strategic planning. For the last five years, he has worked for the Kansas Bar Association as director of lawyer referral.

He says that background makes him the most qualified for secretary of state. With current Secretary of State Kris Kobach running for governor, a total of six people are vying for the office: four other Republicans — Randy Duncan, a former Saline County commissioner, Army veteran Craig McCullah, and state Reps. Keith Esau and Scott Schwab, both of Olathe — and Democrat Brian McClendon, Lawrence, a software designer and engineer who is a former executive at Google.

“My view is that the secretary of state job is really a management and communication position,” Taylor said in an interview Friday with The Hays Daily News. “It’s not a legislative position. We’ve sort of been treating it that way for the last few years. I’m confident the Legislature can handle policy issues on their own.”

His background also includes working in Europe throughout a decade as regional director of the U.S. government’s Eastern European Public Administration Reform Program, assisting emerging democracies after the fall of the Berlin Wall.

“The Russians still wanted to have some effect on their neighbors,” he said of that time. “They were doing then what they seem to be doing in American now, which is messing around with elections.”

Auditing elections is one method that would help preserve the integrity of the election system and maintain trust in the office of secretary of state, Taylor said.

He said if audits had been used in Kansas, Kobach might have had strong evidence to prove the constitutionality of Kansas proof of citizenship law for voter registration. A federal judge in June overturned the state’s law following a lawsuit from the American Civil Liberties Union in which Kobach represented himself.

Taylor said part of the judge’s decision was from the lack of a strong evidentiary record. He discussed that, along with other election issues, with the HDN in a question-and-answer session. His comments are presented here, edited for clarity.

On the Fish v Kobach decision

The strongest evidence, such as it was, were 29 people who’d gone to the DMV to get a drivers license. They were non-citizens. They were presenting evidence they had the right to get a drivers license but they were not, at that point, presenting proof of citizenship

The DMV people, under the federal “motor voter” law, is obliged under that law and the federal funds that come with it, is obliged to allow them to register to vote. Which they did. There were 28 of them that the secretary was not able to say these people actually voted or even tried to vote. Somebody put that form in front of them, they probably signed it because they thought, ‘I’m supposed to be signing all this stuff.’ They didn't go out and vote

He took those 29, or one of his expert witnesses, and extrapolated that information to tell the judge it looks like based on that, as many as 2 percent of eligible voters in the state voted illegally because they were non-citizens. That's 36,000 people. Now there's absolutely no evidence of that other than that extrapolation of 28 to 36,000.

Had they audited in ’14 and in ’16 and actually found people who truly had voted but provided the phony passport, the phony birth certificate, and we'd gone back and audited them, and they couldn't produce a real passport or certificate, then you could have made a case that the proof of citizenship law was constitutional, was valid.

But there's a constitutional principle that you just can't pass a law that has no relationship to what you're trying to accomplish. You're trying to keep non-citizens from voting, fine, but this law isn't going to do it based on the record. So that's why they tossed it out. I don't think they're going to win on appeal.

Now we're back to oath of citizenship under penalty of perjury, a person has to prove they’re a citizen, provide that assurance. That's why I think it’s most important for those people who are absolutely convinced there is some sort skullduggery going on.

To create trust or maintain trust and confidence in the system, we need to audit the eligibility of voters on a post-election basis and be able to go in and assure people with a statistically valid sample size. Twenty-eight people was not a statistically valid sample size on 1,800,000 registrants. So you'd have to have 1,800. That might be statistically valid. if you found 10 percent of those or 5 percent of those were non-citizens, that might have been a case the judge would have been able to say there was a relationship between the law and the purpose.

What would an auditing process be like, ideally?

Ideally, it would be like what they do with statistical polls, survey stuff you get for an election, or the IRS. The IRS doesn't audit everyone in America every year but they audit a fair number of people. What's' the right sample size? As I say, I'm not sure, 1,800 isn't a lot but it would be a lot more than 28.

You can hire companies who do that sort of work and who can give you a statistically valid sample and can say we've looked at this many different people and have them actually have to produce their birth certificate. Not everybody, but just like an IRS audit, if you happen to be the unlucky person who gets audited, then you're going to have to show your paperwork. Then the IRS is able to look at that and say, “We need to shore up our systems because this many people gave us phony data,” or conversely we find out of this statistically valid sample size everybody was clean, everybody gave us valid birth certificates. That ought to give us confidence in the system. 

Do you have any ideas of how much that would cost?

No, not really. I haven’t investigated. I do know it’s being done, post-election audits of eligibility are being done in other states in the United States. So one of the things I would do is research at a more detailed level what other states have done. Sample size would be one issue, and then cost is generally going to be associated with sample size, and then try to apply that to Kansas.

I think whatever the cost is, I think it's the kind of thing ... elections are the foundation of democracy. We need to make sure there's trust and confidence of people in voting, otherwise we're going to create a downward spiral where people say there's not much point in voting. We can't trust whether this is a good system or a bad system. The suspicion among some people that non-citizens are voting, I don't personally believe that, but I can't argue with people against that because it’s all just opinion and speculation. There's not the data to back it up. Data would help a lot. It would put some of those fears to rest I think, or if people are right about that, it might tell us we've got some work to do to make sure that maybe we do need a proof of citizenship law that actually has some teeth in it. But we kind of did it backwards in my view. We passed the law and didn't find out about the data, instead of finding out the data and then passing a law that's related to that.

You said that when you were in Europe, you helped countries with interference in their elections. We are hearing a lot about that now in the U.S. Obviously technology has changed since then, but what do you think you learned then that could help today with that issue?

One of the things is to acknowledge our vulnerability. One of my competitors here in this race is basically saying, ‘Well we passed a proof of citizenship law, we've moved the elections from spring to the fall, now we can just breathe, everything’s fine.’

Were it only to be so. I think that's great, but I think that's a product of hope over experience. We should be more proactive, I think, particularly on constantly verifying all systems’ security.

As secretary of administration, I was in charge of the IT department for the state. Not the secretary of state, but the entire state. One of the things I learned in that job, coupled with my work in Europe, is that systems are always vulnerable. Most of us knew this intuitively. Equifax, the credit people, got hacked, Target got hacked, Saks got hacked, Sony. There are a lot of bad actors out there who are attempting to do something with people's personal information, and the only part that the secretary of state plays in that is we have in the database birthdates, addresses, Social Security numbers, identifying information, and that needs to be kept secure. 

I don't think we're taking that nearly as seriously. We've spent more time worrying about the 12 people who voted in Colorado and Nebraska in the last seven years, and we've prosecuted them. I'm not saying we shouldn’t, because obviously, nobody should vote in two states, but that's minor, those 12 people, compared with the potential at least for the Russians or somebody to get into our systems and create chaos.

They did it in 21 states, the record's pretty clear, in the 2016 election. They didn't necessarily change any votes, but that's very much similar to what we saw in the late ’90s and early 2000s in eastern Europe. They were probing. They were trying to find vulnerabilities and when they found some and found nobody was minding the store, then they could actually go in there and change things. 

That's my fear for ’18, for 2020, is that this probing thing was just a trial run and now we're getting serious. I think some people are too complacent about that. It's not exactly a head-in-the-sand kind of thing. It is hard, no question, to say we need to spend the time and resources and money to prevent these kinds of things, but as they say, an ounce of prevention is worth a pound of cure, and I think if we don't do that, we're asking for trouble. 

Ellis County, like other counties, has just been through its budgeting process and there was discussion of needing to replace its election machines. What kind of things do counties need to be looking at when they're buying new election equipment? Do they need to look at security and how to protect the data and the voting information?

The counties that don't have voter-verified paper ballots, we need to figure out a way to ensure that every county in the state has an ability to be audited, has an audit trail. If you just go and vote on the machine and it doesn't spit out anything that guarantees that vote was counted, you can't come back in an audit, there's nothing to audit. For those counties, I think we need to fix that. For other counties though, the biggest weakness though is not really in the machines, the biggest weakness is in the registration database that the state maintains.

I understand the lifecycle of some of these machines is about 10 years and the cost of some of these machines is, (Ellis County Clerk) Donna (Maskus) was just telling me $350,000 or something. With the property tax lid the state has, there needs to be some sort of acknowledgement on the part of the state that there has to be a cost-sharing kind of arrangement between the county and the state that would ensure the property tax payer is not the one who the entire burden is put on here because property tax here at ad valorem levels around the state are wildly different. Johnson County's ability to generate on one mill, the amount of money is going to be way over what it is in Ellis County, for example.

If we really want a system that is unified and secure throughout the entire state, we've got to figure out a way for everybody to participate in that, not just some people. If the machines are wearing out or you can’t get parts for them and you're going to have to upgrade, there needs to be a way to figure out how to make that work for everybody. I don't know the details of that. I think there's got to be an acknowledgment, though, that we want the foundations of democracy to be protected, and it's not a cost as much as it is an investment to make that expenditure. I don't think there is really a choice.

Anything else you'd like to mention?

Maybe this follows on from what we were saying. I mentioned trust and confidence a couple of times, but that downward spiral, in non-president or presidential years in Kansas the average is about 48, 49 percent of eligible voters now voting. That's not good. and I would want to work, in addition to some of the things we've talked about here, work with civic groups, community groups, others on an ongoing basis to make the public aware of rules and deadlines and the importance of voting.

I’ve run into a number of people in the last couple of months who didn't realize until too late they couldn’t change their party affiliation after June 1, which is a comparatively new law. It's been in effect two or three years. The media I think does a pretty good job, the governments themselves, the secretary of state’s office does a good job right before elections telling people about deadlines and rules and stuff like that, but not on an ongoing basis. 

I think whether it's the school or community or civic groups, I see the secretary of state’s office as a potential convener of those entities to try and encourage understanding and awareness. You can't put a gun to a person's head, you wouldn’t want to do that, and make them vote. That's the way they used to do it in Russia is say, “You will vote.” But by the same token, a lot of the misunderstandings, the lack of confidence and trust in the system is fundamentally a lack of understanding. I think we can do better in terms of making people aware of what their rights as well as their responsibilities are.