BRAINZ!... Zombie Movies and War, An Odd Correlation

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 20^th 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?