Survivorship Bias Applied to Movies

The NYT's Catherine Rampell presents an interesting graphic that compares movies by their return of their production budget.

The problem is that this chart has massive survivorship bias. These numbers were taken from a data set:

"...this data set, which includes movies made from 2003 to 2012 with at least $2 million in domestic box office returns."

Eric Falkenstein's The Missing Risk Premium references a study by Art Devany which shows us that while R-rated movies have high average gross returns but their risk is also much higher. Below is a chart of movie gross return and volatility from 1985 through 1996.

Chart from DeVany and Walls (1999)

Given this information, the original chart is about as accurate as the following study: "I'm going to look at the return of micro cap stocks compared to other stocks, but I'm only looking at the ones that didn't go bankrupt."  Even though we are excluding all bankrupt stocks it's a forgone conclusion that the returns of the micro cap stocks will be significantly higher than any other type of stock. 

The original movie chart isn't devoid of information - it does suggest that globalization is friendlier to NC-17 movies than movies of other ratings.  We've known for a while that the consumption of blue movies is popular in the developing world and this is more evidence.