Election Day has arrived, and as Americans head to the polls for the last day of voting, the country and the world are looking back on an election cycle defined, in large part, by the spread of viral and pervasive misinformation and disinformation.

Leveraging Zignal’s Narrative Intelligence Cloud and metadata associated with digital media postings, we broke out which states saw higher amounts of misinformation narratives, controlled by population and in comparison to their peers. The results show a system that agencies can harness to identify the biggest sources of misinformation, and prioritize them accordingly for counter-operations.

Zignal logo for line break

Key Findings

  • Washington, D.C. has the highest concentration of mentions of misinformation narratives relative to its population.
    • This may be because of its large population of pundits, outlets, and other figures whose main interest is to discuss politics, and that politics dominates a large portion of the social and commercial life in the city.
  • Discussing misinformation narratives is a bipartisan affair. States with a historical propensity to vote both for liberal and conservative candidates share misinformation narratives on an equal basis.
  • Likewise, there are no regional standouts. States that share an above-average amount of misinformation narratives come from all parts of the country – as do states that do not engage in these conversations.
  • States with large populations do not see a dilution of the proportion of misinformation narratives, but are either near or above the average for all states. This shows that even amongst our most populated states, misinformation is being discussed at a very high rate.

How can agencies and contractors use this information?

You’ll see that the information below doesn’t just indicate the states where misinformation narratives are most likely to originate. They also:

  • Show the concentration of mentions of misinformation narratives relative to each state’s population.
  • Create a ranking system for this information so that states with the highest concentration can be quickly identified.

For agencies and contractors charged with rooting out political misinformation, mapping out misinformation narratives in this fashion is a great way to target your efforts. It enables you to see:

  • Where the highest raw volume of misinformation is coming from, so that counter-operations can be directed to areas of highest potential impact.
  • What regions have a disproportionately high output of misinformation narratives on a per-capita basis, so you can investigate root causes and proactively counter future misinformation campaigns.

Trying to figure out the best tool for rooting out misinformation in the public sphere? Check out our Narrative Intelligence Solutions Buyer’s Guide.

Overall Mentions of Misinformation Narratives

Above are the total number of media mentions, broken down by state, that reference one of the hundreds of misinformation narratives Zignal Labs has been tracking, in conjunction with a major candidate in the 2020 election (all Presidential and Vice Presidential candidates, along with other influencers related to the November 3rd election, including notable Senate and House Candidates). Location is determined by the metadata attached to media mentions, which can include MSA information for broadcast television, the city of a local newspaper, location data attached to social media postings, and more.

At first glance, when looking at any geographical breakdown of digital media, the list can appear to be simply a list of states sorted by population. However, when we compare these state mentions against each other, certain contrasts and insights come to light.

For this exercise, Zignal Labs ranked each of the locations by the volume of total mentions, and compared these mentions with their rank of total population (source: Wikipedia). Subtracting mention rank by population rank gives us a differentiation number. Zero means a state’s population matches its misinformation mention rank. A negative number indicates a disproportionately high amount of mentions of misinformation narratives relative to its population rank, and vice versa.

Zignal also compared the mentions of misinformation narratives against the state population for an overall percentage. The average percentage of misinformation narratives mentions to total population for each state is 8%.

Comparing States

When comparing states against each other by rank differentiation (subtracting population rank from misinformation rank), we are able to make an easy comparison of which states are more likely than others to discuss misinformation narratives.

Within this view, Washington, D.C. is far more likely than its peers to discuss topics related to misinformation narratives, while its immediate neighbors of Virginia and Maryland are among the least likely.

Notably, there is no discernible regional or partisan split among these states. All regions – and states that reliably vote on both sides of the partisan divide – are present in each of these metrics.

Percentage Breakdown

As a percentage of population, Washington D.C.’s mentions of misinformation narratives is staggering. West Virginia and Oregon (which both ranked highly in the differentiation metric) are among the states with the highest rate of misinformation discussions by population.

Keeping in mind that the average percentage for all states is 8%, some states stand out. Florida and Texas did not rank in the differentiation metric, as their total mentions of misinformation narratives rank lined up with their population rank. However, they both produced a higher than average amount of mentions of misinformation narratives as a proportion of population, which is even more significant given how large their state populations are.

The 2020 election is like no other in plenty of ways, from the pandemic-driven expansion of early voting to the level of tension among the voting public and beyond. The proliferation of political misinformation is just one of many things that have made this election stand out. But just like better understanding the novel coronavirus can help us fight the pandemic, better understanding the political misinformation landscape can help us ensure more sound elections.

Methodology and Scope

This analysis was conducted within Zignal’s “All Candidates” profile (1.8B Mentions) and filtered based on keywords related to various misinformation narratives identified throughout the year (703,570 Mentions).

  • Date Range: February 1st, 2020 – October 29, 2020
  • Please note that Zignal Labs does not determine what is true or false. Rather, Zignal looks at stories and themes that surface in the media and provides data regarding origin, spread, degree of automation involved in the spread, and the like.
  • Location data is contingent on data sources in which end users have opted into location sharing, presenting a representative sample of each geographic region.