The results of the 2020 election might have taken weeks to resolve, but now there’s fresh evidence the balloting was kosher. That’s according to detailed analyses conducted by the MITRE Corporation’s National Election Security Lab. With how they did it, MITRE’s director of cyber integration, Emily Frye, joined Federal Drive with Tom Temin.
Federal News Network's Cyber Industry Exchange: Where does cybersecurity end and physical security begin for federal agencies? Find out at Federal News Network’s Industry Exchange: Cyber.
Tom Temin: Ms. Frye, good to have you on
Emily Frye: Good morning Tom, thanks for having me.
Tom Temin: And let’s set the background here. Why did MITRE undertake this particular study of some of the balloting that happened in the contested states, in particular what caused you to take a look at this I guess after the courts had already made their decisions?
Emily Frye: Well we actually started to look at election related issues when we started the SQUINT program, which was a way of collecting misinformation all through 2020, and we could see in the signs that were coming in from the SQUINT program all kinds of indicators that this could be an election that had an unusual level of concern, tension, maybe contention and so we decided to really try to step in and offer MITRE’s data analytics capabilities to the nation in a time of need.
Tom Temin: Alright. And tell us the methodology because mail in ballots happened in different ways in different jurisdictions and it would seem to be a hard thing to reconstruct afterwards — so how did you go about determining the accuracy and efficacy of that mail in balloting.
Emily Frye: That is a great question. So the main thing that we were able to do was simply watch the news and figure out where are these allegations of mistrust, concern, fraud coming from and really that was our jumping off point. We watched for the items of concern in the news and this is the important part, we took those questions and we began to ask ourselves where is there data and evidence that would actually allow us to address or answer a question that we have seen in the media. And there were many such questions where there was simply no data and no evidence, so we did not address those questions because we were only going to do what if we had data so the evaluations that are summarized in the study that was published are the places where we found good quality data that we could really use and assess the issue. There are six such examples
Tom Temin: And where were those?
Emily Frye: So this is interesting, it is a range across the country. We’ve got one in Butler County, Pennsylvania, we’ve got one case study in Antrim County, Michigan and an awful lot of case studies around Georgia, in the end that was a coincidence but it happened to be very useful because there was a lot of attention on Georgia.
Tom Temin: Sure, and what were the data sets that you actually ran through, and then we’ll talk about some of the algorithms or what the research methodology was — but what were the data sets themselves?
Emily Frye: We had access to a number of data sets that were being published by the election officials in the different jurisdictions by the state election directors and the Secretaries of State. So in addition to that, we were accessing publicly available datasets from different news organizations. Each analysis is a little bit different, we had easiest access to the data in Georgia and I would say hats off to them for remarkable transparency, they did a great job.
Tom Temin: So these were the data sets of the counts of the paper ballots, of the mail in ballots?
Emily Frye: So we actually looked beyond mailing ballots to other sets of allegations, and let me give an example if i may. One of the issues that we looked at was whether or not there was an anomaly in returning mail in ballots in Butler County, Pennsylvania. So early on in the process of pre-November 3rd voting, we could see that there was a county that was not returning mail in ballots based upon simply the numbers that were collected of the returned ballots and that data is collected by the state of Pennsylvania, so we were able to take a look at that. So that raised the question for us and presented the question for us why would there be this one county where there are no ballots being returned, this doesn’t make any sense. Is this is this a suspicious item and when we looked into it we realized that as a matter of fact the Postal Service had misplaced a large number of ballots and they simply had not been delivered to the residents around the county. So you can see we’re using a number of different sets of data. We’re using data from the Secretary of State’s office, we’re using data from the Postal Service, we’re using data from the news so that analysis is actually tailored to where we can find the data. What we were able to find in that case was at one point in the process, the Postal Service caught up with its deliveries and at that point in time, the returns of mailed ballots then began to catch up with the norm. And so by the time we reached election day, there was no longer an anomaly.
Tom Temin: And with respect to what some alleged were unreal results from those mail in ballots, they were all for Joe Biden or something, that type of thing. Were you able to verify or discount that particular phenomenon? That it was out of band for what might be expected vis-à-vis ballots cast the standard way?
Emily Frye: That’s a great question. So there are a couple of different instances in the study where we’re looking at the difference between what happens when you have mail in ballots, and what happens when you really just have people voting on election day. So we looked at the difference between recorded votes and types of party line affiliations that showed up in 2016, as opposed to 2020. And we really saw, and this has been reported by others so I feel especially confident in saying this mail in voting, it does increase the turnout. And we want that as a nation, we want our voters to be engaged. But while increasing the turnout, it does not have an effect on the party that wins or loses. That is not a correlation that we really see.
Tom Temin: Interesting. So adding all of this up, do you see lessons learned for local jurisdictions that operate elections? And is there any lesson to be learned by the federal government, which in some cases, does set policy it’s not supposed to in other instances, and so on? So what can we all take away from this for the next time around?
Want to stay up to date with the latest federal news and information from all your devices? Download the revamped Federal News Network app
Emily Frye: Sure. So let’s parse that apart for just a minute, I’m going to go back to the state of Georgia, we were able in the state of Georgia to find quality and regularity of data that was really not available in the majority of states. So our recommendation to states is get the data collection methodologies in place so that there is maximum transparency in the run up to elections, during elections, after elections, and make this data available to the independent researchers who are trying to investigate integrity of the elections. And let me give you an example, we were able to see in the state of Georgia that they were collecting data down to the individual level, not at the state level, not just at the county level, but down to the individual as to whether they had returned a ballot and whether that ballot was accepted or rejected, and the value of that data in determining the integrity of an election cannot be overstated.
Tom Temin: And do they do that with some kind of a barcode system or how are they able to do that?
Emily Frye: Yes, so they actually do have a bar code system and a number of other very human intensive methodologies that they’re using to track every ballot. So the point that I am taking away from this, and I really hope that this is shared as broadly as possible, there is a role here for the states to be as thorough in data collection as they can be. And there’s a role for the federal government in really enabling and empowering states to have the resources that they need on a sustained basis to conduct both the cybersecurity and technical assessments that they need to collect, but also to really put in place those processes that enable them to communicate and support integrity and the perceptions of integrity with the public.
Tom Temin: And then there’s also the issue of chain of custody of ballots. And that’s something you can really never let your guard down on, is it? Including the Postal Service.
Emily Frye: Yeah, that’s really true. A chain of custody is very important. One of the things that different states have been doing is the barcode system so that it’s in fact possible to track where these things into the system, where they move through the system, where they end up in the system.
Tom Temin: Alright, a little technology and a little human integrity, and maybe we’ll make it through I guess. Emily Frye is director of cyber integration at the MITRE Corporation. Thanks so much for joining me.
Emily Frye: Thank you so much, Tom. It’s been a pleasure.