Where is disinformation the biggest threat
Modelling Risks to Overturning an Election with Disinformation
Voting in the midterm elections ends tomorrow, meaning we’ll begin to quickly learn who won, who lost, and what races are going to be contested. There are a number of risks to elections and voting, including risks of voter disenfranchisement, fraud, or further exacerbating the loss of trust in American elections. We’ve discussed how disinformation can contribute to undermining elections. Here, we examine the greatest possible risk — that disinformation leads to changing the result of an election.
The 2020 Disinformation Playbook
Disinformation has long changed election outcomes, whether by undergirding laws limiting who can vote to mobilizing widespread voter intimidation under Jim Crow. The 2020 presidential election provides us with the freshest playbook for how disinformation about American elections emerges, gets used by activists to mobilize politicians who then use the apparatus of the state to influence election outcomes.
In previous posts, we’ve discussed how disinformation about the midterms might get created and what can be done to address it. The Election Integrity Partnership has also released it’s guidance for 2022 and is a must-read for anyone addressing disinformation in real time.
In this post, we want to step back to ask which races are most at risk for disinformation to impact the final result. The goal is to help identify the greatest potential threat in order to prioritize resources where they are most likely needed. This post also serves as a pre-election pre-registration of our model of risk for an overturned election. With few elections between now and 2024, we have few opportunities left to pre-register any hypothesis.
Model of Election Integrity Risk
How might an election be overturned?
The greatest risk to an election is to have a valid result overturned such that the candidate or party winning the most votes does not win the seat. In 2020, we saw this happening three ways:
First through executive intervention. This is what Donald Trump tried to do in 2020 by having Pence reject votes on January 6th. It’s also what happened this summer in Otero County, NM when the county canvassing board refused to certify their own election despite a lack of evidence of improper voting.
Second, legislatures may try to intervene by rejecting a vote and installing a losing candidate. This is what the “fake electors” plot tried to do. It is also the subject of the Supreme Court’s upcoming Harper v. Moore case, in which the state legislature of North Carolina is asserting unilateral control over election procedure.
Finally, the judiciary can intervene to overturn election results. This is what we saw at the state level in 2020 when the constitutionality of many votes were challenged in court because of the way they were cast. This type of overturning is tricky because courts are the valid mediator for interpreting laws. If courts rule some votes violate the law and that ruling changes results, then it generally cannot be considered overturning an election. By definition, the only way the judiciary can overturn an election would be through intentional or gross negligence. However, the persistent election denial by Michael Gableman, the former State Supreme Court Justice in Wisconsin, suggests it’s not unthinkable.
[CUT?]There are a couple of related outcomes that may occur but are not necessarily overturning an election. One outcome would be having an election be held again. Rerunning elections generally happens when systemic fraud is found as happened in NC’s 9th district in 2019. But, we can see it happening erroneously (i.e. in the absence of real fraud) if disinformation convinces a state to rerun the election.
Model of Risk for an Overturned Election
Under what conditions might disinformation lead an election to be overturned? Given that disinformation is factually unconstrained, it can be used to make any claim and target any race. In other words, disinformation by itself does not overturn elections. It’s just a tool. Instead, it works through a political system, the activists, officials, journalists, and politicians. It’s the vulnerability of the political system that puts an election at risk.
Outcome Closeness: There’s no reason close elections should be targeted for overturning. Trump claimed fraud happened in every state. But, he applied pressure to only a small handful of the closest states. Why does closeness matter? It may be surprise. Candidates and people may be more willing to believe claims of fraud in competitive races. It may also be a sense of political support. Whatever the reason, uncompetitive races were not targeted for overturning by disinformation in 2020.
Republican Loss: At present, disinformation about election integrity is more common among republican voters. (This may change if disinformation about disenfranchisement emerges on the left, but for now there is little evidence to support it.) So, Republican voters and elected officials are more likely to claim an election was stolen and no one claims an election was stolen if they’re declared the winner.
Republican Control of Legislative or Executive Branch: As Republican officials and party members are more receptive to election denial disinformation, they are more likely to act on it. At the state level, we’ve seen this in republican legislators in Arizona and Wisconsin endorsing the decertification of the 2020 election. The more levers of power republicans control, the greater the likelihood that they would use one or more levers to overturn the election. However the legal path to overturning an election based on disinformation, that is to say, unfounded claims of election fraud, is still very untested. 2022 may be the year we see more paths tried out.
Grassroots Mobilization: An essential component of the 2020 Stop the Steal movement and subsequent denialism has been the mobilization of grassroots deniers against elections officials and republican representatives . Election deniers have been notoriously aggressive in pressuring (or harassing) politicians, election officials, law enforcement, and journalists to overturn elections. Empirically the big question is how much of a role a grassroots mobilization will play and who will be targeted. Historically, the mobilization has found its most receptive audience in republican politicians and sheriffs but not elections officials, democrats, or judges. More likely than not, grassroots mobilization won’t overturn elections. Instead, it will enhance the effect of the first three, magnifying pressures to overturn a close republican loss in a state controlled by republicans.
Predictive Wildcards:
Legal Wiggle Room: The big unknown is how successful such a movement can actually be. How many pathways to overturning an election in the absence of substantive evidence of fraud are there really? Can legislatures unilaterally reject or rerun an election? Can governors stop counting votes? Are there judges out there willing to reject elections based on unsubstantiated claims and if someone does, what is the chain of appeals?
Disinformation Elite Focus: Overturning elections takes political will. Elites with large followings like Donald Trump and Tucker Carlson can create substantial political pressure if they focus on a particular race. The question is whether any disinformation claims can garner the attention of these elites. I believe elites respond to opportunities, so they would likely amplify disinformation about at-risk elections. But, as elites, they would also be focusing on high-profile elections like gubernatorial or senatorial elections rather than lower profile house races or secretaries of state. Expect this to go out the window if the house or senate end up being at play. If control of the house comes down to five seats and Republicans are on the losing end, then elites will be more likely to step in.
Who is most at Risk?
Based on the model above, the states at highest risk for an election being overturned are those with close gubernatorial and senate races that are tossup or lean democratic but in states where republicans control one or more branches of government, i.e. Pennsylvania (1 close race), Georgia (1 close race), Arizona (2 close races), New Hampshire (1 race), Wisconsin (2 close races) and Oklahoma (1 close race). Nevada, Oregon, and New Mexico have close races but are run by democratic supermajorities. While they may be targets of disinformation, the risk that one of their elections will be overturned is likely small. Each election is likely to be independent, meaning if the republican loses the close election in one race but not the other, you will likely still see a higher likelihood for overturning the election.
Caveats & Limitations
2022 is the next iteration on election disinformation and one of the last trials we get before 2024. But, every election has its own idiosyncrasies and there are some basic things that could change the calculus.
A Republican Wave: This is expected to be a good election for Republicans. If they win the races they’re leading and sweep the tossups, there won’t be a reason to overturn the election. While this is a good thing for democracy, it means we’ll likely learn little about more contentious elections.
A Blue Wave: On the other hand, an unexpected blue wave would bring new states into play like Texas, Florida, Ohio, North Carolina, and Alaska. This would provide more evidence for whether and how elections may be challenged in different states and, given a broad win by democrats, may better approximate a presidential election where many states contribute to the overall outcome.
2024 will be on steroids: Midterms are quiet affairs typically. And, Trump is not on the ballot. Compared to 2024, the disinformation this year may be relatively mild.
Changing State Laws and Officials for 2024: A big part of the 2020 stop the steal movement was legal challenges to pandemic-era voting laws and procedures. Now, these issues have become more settled after two years of litigation and this will be truer in 2024. That said, states may pass new laws making it easier or more difficult to overturn an election. And, election deniers may win key positions like secretaries of state. These will shape the opportunity space this year in ways that are different from 2020 and will be different from 2024.
Conclusion
At baseline, having an election overturned is likely a low probability event. I’m not sure it has happened since the Civil Rights Act of 1965. But, there are people who are actively using disinformation to overturn elections and a substantial part of the population who believes American elections are fraudulent.
Monitoring disinformation is about identifying what claims are being made about whom. They are the arguments being made to overturn an election. This risk report is about prioritization - what elections should we keep more of an eye on? What disinformation poses a greater risk to voting rights?
We’ll be watching the returns starting tomorrow and the disinformation being put out as it evolves. We’ll continue to put out reports to help provide guidance to combat election disinformation and ensure free and fair elections