I have to admit, when I first read the post “War and Social
Upheaval Causes Spikes in Zombie Movie Production” , I was a bit skeptical
about the conclusion that was drawn dealing with the correlation between
war and the number of zombie movies produced in a given year. So I collected
data and ran a model. As it turns out, there is a statistically significant
relationship between the number of zombie movies produced and whether a war was
fought within a two year time period.
More after the jump…
The potential causal explanations behind this observed
correlation are endless. The authors of the Freakonomics blog and their readers
offer some really interesting explanations as to why this relationship exists. Perhaps
zombies reflect part of who we are and zombie movies help us come to
terms with ourselves, or perhaps zombies symbolize the hope that we can
overcome great obstacles like death.
If I had to offer a causal explanation as to why this
relationship exists, I would say that people see the enemy in a war as not
human, but something less than human, scary, uncontrollable, and deadly, like a
zombie. So maybe people are more likely to relate to zombie movies during war
times because they are dealing with something similar themselves. I would also
say that maybe this is just a correlation with no causal force behind it.
For those who are interested, here is a brief explanation of
the model I ran and the results:
The dependent variable I used in the model was the number of
zombie movies produced in the US or Europe within a given year. I debated
whether to only examine the relationship between wars and zombie movies within
a given state. I decided not to because given the close political, economic,
and cultural relationships between European states and the US, it is safer to
assume that the foreign policies of a given state in this sample have some
impact within the other states in the sample. For this variable, I used the
data the author of the post provided.
In the model, I included two independent variables. First
was whether a war was fought in the US or Europe within the last two years. I consulted the Correlates of War data
to determine which years wars (as defined as 1000 or more battle deaths) were fought in either the US or Europe. I lagged
the variable because if a war leads to a movie being produced, it will take two
years for the movie to be finished and hence the effect of war will lag. I also
included time as a variable to control for the fact that more movies are
produced today than they were in the early 20th century.
I ran a Poisson model because like war and death by mule
kicks, zombie movies are a rare event. Here
are the results:
The results indicate that if there was a war in Europe and the US within a two year time period during the 20th century, then the number of zombie movies would increase…. correlation or causation?