Thread by @ProfPaulPoast

1/ Easter, like many religious holidays, has been tied to peace & conflict.

We explored the link b/w religious holidays and violence in my Quantitative Security course.

👇is an Easter morning summary.

2/ There is great work exploring the link b/w religious beliefs and violence, such as this @Journal_IS piece by @monicaduffytoft

mitpressjournals.org/doi/abs/10.116…

3/ But a newer line of work directly explores how religious holidays alter patterns of violence.

Examples include Ron Hassner's @SecStudies_Jrnl piece exploring the relationship between "Sacred Time" and conflict initiation...

tandfonline.com/doi/abs/10.108…

4/ ... and a recent APSR piece by @WorldReese, Keven Ruby, and Robert Pape on Islamic Holidays and violence patterns in Afghanistan, Iraq, and Pakistan

cambridge.org/core/journals/…

5/ So the idea is that major religious holidays influence violence patterns. But much of this research looks at post-1945 cases of conflict & focuses on regions with Muslim majorities.

Examples include insurgencies in Iraq or, perhaps most famously, the 1973 "Yom Kippur War"

6/ Does the "Sacred Time" hypothesis apply outside those contexts?

We explored that question looking at the American Civil War

7/ While there are a host of ways that we could measure violence in this war, we chose to look at instances of battle initiation.

A tremendous resource on each battle fought during the American Civil war is the @ABPPNPS of the @NatlParkService:

nps.gov/abpp/battles/b…

8/ They catalogue, by US State, the Civil War battlefields. Here is the list for Alabama

9/ Clicking on a battle provides all sorts of details about the battle. Here is the entry for the Battle of Fort Blakely

10/ Other scholars have used these data to create a county-level map of battle locations throughout the war

11/ Since we are interested in testing the "Sacred Time" claim, we decided to look at the start date of the battles.

We first entered these into Excel. Here are the first several battle-initiation dates

12/ We then took these data over to #Stata and created the below figure.

Using data over the course of the war (April 1861 to April 1865), this figure shows the number of total battle onsets by month:

13/ A few things jumped out to us:

- First, you can see that spring (April, May, and June) is definitely "fighting season". No surprise there!

- Second, notice that early spring (March) appears to be as "disputatious" as December. That stood out to us for two reasons...🤔

14/ For one thing, you would think that the weather isn't all that great in December (definitely NOT fighting season -- except for maybe snowballs).

15/ For another, isn't there a MAJOR Christian Holiday in December?

16/ Given the cultural dominance of (largely Protestant) Christianity in the US at that time, we looked a little more closely at the patterns of battle onset in December.

Here is what we found:

17/ As you can see, NO battles were initiated during the week of or ON Christmas. It does appear that this religious holiday was a time of restraint. A bit of Christmas Peace!

18/ But wait a minute! While it seems that Christmas Day was a time of Restraint, notice the day after?

The spike suggests that while the sides did lay low for Christmas day, it's back to business the day after!

19/ What about Easter, the other major Christian Holiday (and, as someone raised Catholic, THE major Christian Holiday)?

20/ Easter is a bit tricky b/c, unlike Christmas (which is ALWAYS December 25), Easter jumps around 🤔🤷‍♂️

en.wikipedia.org/wiki/Computus

21/ The American Civil War witnessed three Easters:

- April 4, 1862
- April 6, 1863
- Marcy 28, 1864

22/ Of course, some might view the seemingly random placement of Easter as an opportunity to create an instrumental variable.

But that makes me nervous

https://twitter.com/ProfPaulPoast/status/1119564726432870401

23/ And three days isn't much information on which to draw an inference. Still, for what it's worth, no battles were initiated on those days.

So perhaps Easter was treated as a "day of restraint" as well 🙏

24/ This exercise allowed students to see what they can learn by doing a "deep dive" into micro-level data.

This lesson was then reinforced by exploring the Rwandan genocide with data from @engagedscholar & Stam:

https://twitter.com/ProfPaulPoast/status/1115946304549609473

25/ In summary, the American Civil War allowed us to see that the idea of "Sacred Time" as a "Day of Restraint" does appear to apply to pre-1945 conflict and outside of Islam.

(end)