Instability risks of the upcoming U.S. election and recommendations for EAs

Written by Juan Cambeiro and Kyle Bogosian. Views are our own.

Disclaimer: We both support Joe Biden for president, and Kyle has done a comprehensive analysis from an EA perspective on why Biden is far preferable to Trump on nearly all EA and EA-adjacent issues. However, our recommendations in this post should still mostly apply even if you do not support Biden.

Summary

A recent forecasting session by Metaculus of top Metaculus forecasters and political scientists with expertise in U.S. electoral politics has found significant risks to the upcoming U.S. presidential election, including on issues relating to effective electoral administration, potential contestation of results, likelihood of civil unrest, and smoothness of any potential transition. We discuss the implications this has on the near-term stability of the U.S. and steps that American EAs can take over the coming days and weeks to help mitigate these risks. In particular, we recommend: voting, encouraging others to vote, not engaging in protest activity or poll watching on Election Day, exercising patience for results, being careful not to spread disinformation, and supporting sustained post-election peaceful protests in the event that the integrity of the election is seriously threatened.

Metaculus’s 2020 election risks survey

Metaculus and Rachel Aicher, a political scientist at CUNY Hunter College, recently organized an expert forecasting session on risks to and of the 2020 U.S. election. This session took place from 8 October to 15 October 2020 and involved a total of 17 experts — 8 subject matter experts and 9 generalist forecasters. Below we highlight findings that are especially relevant to this post. You can find the full report here and you can watch a discussion of the results here.

-Voter intimidation and challenges: Experts’ median prediction for the number of voter intimidation/challenge calls that will be made through the Election Protection nonpartisan voter protection national hotline is 8.4k (80% CI: 1.5k − 18.3k). The number of such calls made during the 2016 election was approximately 4.7k. Experts assigned a 5% probability to over 20k calls being placed, which would be more than four times the number of calls made in 2016.

-Presidential election result known: Experts predicted a median of 8 November 2020 (80% CI: 3 November 2020 − 13 January 2021) for when the presidential election result will be called by the Associated Press.

-Concession by losing candidate: Experts assigned a median of 40% to the probability of President Trump conceding if the election is called by the Associated Press in favor of Democratic nominee Biden. They assigned a median 95% chance to Democratic nominee Biden conceding if the election is called by the Associated Press in favor of President Trump.

-Domestic election interference: Experts’ median prediction for President Trump ordering an agency or personnel under his authority to directly interfere in the vote counting process and for these orders to be successfully carried out is 10%.

-Election-related violence and unrest: Experts predicted a median of 60 deaths (80% CI: 0 − 912) occurring due to election-related violence between 20 October 2020 and 20 January 2021. Experts assigned an 8% probability to over 1,000 deaths. Experts’ median prediction for President Trump invoking the Insurrection Act between 3 November 2020 and 20 January 2021 is 10%.

-Confusion as to who will be the next president on Inauguration Day: Experts assigned a 2% probability to it being unclear who is exercising the powers of the presidency after 12PM EST on 20 January 2021.

What to do for this election

Vote: At this point, either vote in-person or drop off your ballot at an official ballot box. Consult your state’s election board website or FiveThirtyEight’s state-by-state guide to voting. We think it is particularly important for you to vote if you live in a swing state — especially Pennsylvania, Wisconsin, Nevada, New Hampshire, or Michigan. In the aforementioned states, the probability of one vote changing the outcome is greater than 1 in 15 million. For further reading on this topic we recommend Rob Wiblin’s recent article, where he discusses the high expected impact of voting in U.S. presidential elections.

Encourage people in your social network to vote: Research shows that encouraging others in your social network to vote is effective. A randomized get-out-the-vote study by Turnout Nation and political scientists Donald Green and Oliver McClellan found that “...turnout rates are 13.2 percentage points higher in the randomly assigned treatment group than the randomly assigned control group, the largest intent-to-treat effect documented by an experimental GOTV study over the past two decades.” It may be a little awkward, but it’s probably worth reaching out to your family members, friends, and acquaintances to make sure they vote!

Do not engage in protest activity or poll watching on Election Day: We strongly recommend against people attending protests or participating in poll watching on Election Day, as these activities are likely to be seen by some people — especially those from traditionally disenfranchised groups — as voter intimidation. We are moreover confident that there are no real benefits to these activities — we do not think there is any reasonable goal that Election Day protests could aim to achieve and we think poll watching is mostly useless since voter fraud is extremely rare. Moreover, as we’ve indicated, these activities are likely to cause harm by decreasing voter turnout and increasing the probability of volatile confrontations at the polls. The number of voter intimidation calls is already likely to be very high this election relative to 2016 — don’t make it any worse!

Exercise patience while waiting for results: In the event of a Biden landslide, it is likely we will know the winner of the presidential election tomorrow night. In particular, keep an eye on the fast-counting states of Florida, Georgia, Ohio and North Carolina — a Biden victory in any or multiple of these indicates that it is extremely probable he will have won the election. However, there may be a protracted vote counting process if the election is closer and comes down to the slow-counting states of Pennsylvania, Wisconsin, and Michigan. The latter scenario is probably the main reason why the median for when an election result is expected is 8 November. Given the changes in voting procedures due to the COVID-19 pandemic, especially the increase in voting by mail, such delays are entirely within the realm of reasonable expectations and do not mean that anything is wrong. It certainly does not mean that there is widespread voter fraud. We recommend that you exercise patience while waiting for results and that you encourage others in your social network to also be patient (go to bed tomorrow night!). We also advise that you get your results-related information from reputable outlets, in particular ABC News, AP, CNN, CBS News, Decision Desk HQ, Fox News or NBC News — all of which are widely agreed to have nonpartisan and trustworthy decision desks.

Be careful not to spread disinformation: It is likely that there will be active disinformation efforts by domestic and foreign actors that will attempt to sow doubt and confusion. Be skeptical about any such rumors and do not uncritically repeat them. Do not assume that something is true just because a few random blue-checks say so on Twitter — they can make mistakes too. If videos or images circulate purporting to show violent incidents, note that these may have been taken previously from some other context and rebranded. Trust reporting from reputable news networks.

Support post-election peaceful protests if the losing candidate refuses to concede: The current Metaculus prediction for a Trump victory is 7% (a good sanity check is that this is between the probabilities assigned by FiveThirtyEight and The Economist). When this probability is combined with a 40% probability of President Trump conceding if he loses and a 95% chance of Biden conceding if he loses, it translates to a ~44% chance that the losing candidate will concede. The bulk of the remaining 56% chance of the losing candidate refusing to concede is from the scenario in which President Trump refuses to concede. Indeed, he may go beyond this and try to declare victory prematurely or even order direct interference in the vote counting process (median chance of this latter scenario occurring and being carried out is 10%). In any of these scenarios, sustained peaceful protest activity may play an important role in exerting pressure on relevant institutions to recognize the legitimate winner of the election. However, extreme caution on the part of protestors is warranted since there may be significant election-related violence and because President Trump may invoke the Insurrection Act in an attempt to suppress protests.

Why this is important

The democratic status of the United States is not only important for Americans (though Americans are in by far the best position to influence it). The United States is the world’s most powerful country and will probably remain one of the most powerful countries for much of the twenty-first century. It is widely considered to be an important bulwark against authoritarianism and illiberalism, and there is substantial evidence that democracy and liberalism are better for human rights, economic growth, and world peace. We also think that maintaining the global role of a democratic United States may be critical to preventing the permanent lock-in of a totalitarian regime. Therefore, American EAs should help ensure the integrity of this election since there is probably high expected value in doing so.