Election night in America: Do national elections lead to an increase in domestic violence?
By Lucy Morisse-Corsetti (Class of 2027)
When I was a kid, Election Night felt like the Superbowl. We got to order take out, my siblings and I would all be dressed up in matching pajamas, and we would sit with a map and our red and blue crayons waiting for Wolf Blitzer to tell us what state to color in next. It was a fun party, a chance to stay up late, and nothing more.
As an adult, both events have lost some of that glimmer of magic and have gained a bit of fear. As fun as the Super Bowl is, the internet is always sure to remind me of a less-than-fun fact: that domestic violence is often reported to have a notable spike during the game. In general, the tension of a football game, in combination with the expectation of what the event “should” be, leads to an increase of violence towards a domestic partner. When looking at this trend, there is no obvious relationship to the Superbowl as a unique event, rather, game days of local NFL teams are clearly associated with higher rates of family violence. For example, “upset losses” (losses where a home team was predicted to win by 4 or more) lead to about a 10% uptick in police reports of at-home male-on-female intimate partner violence. What if elections — which, in my experience, are an equally stressful event — have a similar effect? What if all of that stress and fear and anger coupled with a reliance on a television broadcast create frighteningly similar outcomes?
According to Cynthia Peacock, an associate professor of Communication Studies, political alignment is “tied to our morals, to our identity, to things we hold very deeply and personally.” However, her research suggests that the tension is less attributed to the difference in ideology itself and more to a couples ability (or lack thereof) to talk about these differences. Different couples will take different approaches to this conflict: some avoid, some try for civil debate, some descend into full blown arguments. The latter of these options is more likely to lead to assault by an intimate partner. In fact, these fights themselves can be considered domestic violence, with the legal parameters varying state by state.
When beginning to investigate the possible connection between domestic violence and national elections, there were several potential outcomes. The first was that there is some statistically significant correlation between the elections and domestic violence. The previously mentioned 10% uptick due to sports games is a high bar to pass, but that doesn’t mean that a lesser (but still distinguishable) effect isn’t possible. However, there is another possible outcome: that there is no connection. Elections might just not have the same emotional grab on people that sports do.
To investigate this question, I analyzed the NYPD’s call logs, which track all 911 calls in the city since January 1st, 2018 and track incident type and date (among a variety of other factors). This is one of the few cities that keeps this kind of data in a readily available format. I sorted these incident counts by weekday, with each year denoted by a different color. I was then able to create a trend line of an average week for each year, which showed that there was a consistent distribution of events: weekends were the peaks for 911 calls related to domestic violence, and Wednesdays were the low points. I then highlighted election nights (which fall on Tuesdays), as well as Super Bowls (which are on Sundays).
While the Super Bowls in this sample were not statistically found to be connected to an increase in domestic violence, this would be expected as previous research has found effects to be localized, and the New York Giants (the local team) have not appeared in the Superbowl since 2011. It’s clear from this graph that election nights do not deviate from the normal trend lines, and I confirmed this with a variety of statistical tests. In fact, the 2018 midterm election sits significantly below its annual trend line, and the 2022 election is approximately in line with its annual trend as well. The 2020 election does sit slightly more above the pink line that indicates the average number of domestic violence events in 2020, but this is nowhere near an outlier. I also looked at a 10 day pre-election or post-election period, which similarly did not stand out from the normal weekly trends shown in this graph — there were peaks on the weekends (Saturdays and Sundays), a dip leading to Wednesday and an increase back to the weekend. Furthermore, the weekends that bookended the election didn’t show any sort of deviation from the average as well.
When pulling out just the Tuesday data, the normality of election day counts is even more obvious. This data doesn’t demonstrate the same direct connection that domestic violence and sports games do, and I would classify these results as falling under the theory that there is no connection. However, this seems like it is a deviation from what one would expect, and there are a variety of explanations as to why this might be happening. The first of these is simple — some people don’t care about elections as much as they do about sports. According to a poll published by Seton Hall University, 28% of males say that they would rather have their favorite team win a championship than their preferred candidate win the 2020 presidential election. While nowhere near a majority, this is a noteworthy statistic in an election that was highly polarizing and prevalent in the media — and perhaps a clear indication that elections might not have a strong emotional pull for men when compared to sports. Furthermore, sports are inherently more violent in nature than elections, where seeing someone getting tackled on screen is perhaps more emotionally stimulating than hearing election results.
However, there is another possible (and potentially simultaneous) factor that could be at play here. In my previous research, I found that there is not a notable difference in what people of different ideologies want out of relationships. Furthermore, a recent study from the University of Michigan found that less than 8% of romantic relationships are between a Republican and a Democrat. On the other hand, NFL interest varies widely by gender, with only 24% of women saying they are avid fans as compared to 51% of men. If a couple feels equally invested in the outcome of the event, and they want the same outcome, the event itself might be less likely to prompt violence. This might mean that it may be more common for men to feel isolated or not supported by their partner during a game as opposed to an election, where their interests are more likely to align. Furthermore, we know that a majority of domestic violence incidents are male-on-female partner violence. While it goes without saying that there is no excuse for any kind of domestic violence, this might provide a more nuanced reasoning as to why a NFL game might be more of a flash point.
There may be room for further investigation into the question, perhaps by looking at a swing state, such as Pennsylvania. However, this initial investigation has found no correlation between the election and domestic violence, and I have hopefully provided some explanation as to why my initial conclusion didn’t play out as expected. At the end of the day, this is of course a good thing: it means that there’s one less day where we need to worry about abnormally high levels of domestic violence. This project’s lack of outcome is, in fact, the preferable result.
If you or anyone you know is in need of help, here are some resources for victims of domestic violence.
National Domestic Violence Hotline | Identifying Domestic Violence | Common Signs of Domestic Violence
