Did Star Wars ruin movies?

December 29, 2017 | EdnaMode

[DoodleBob]

I agree with BwanaNdge: Serious questions call for serious analysis.

I took the data Edna referenced (for the top 300 grossing films of all time, with grosses including DVD re-releases which increases Disney films meaningfully). This time series had gross revenue that was adjusted for inflation, which became my dependent variable (I also modeled the natural log of the adjusted revenue which slightly strengthened the regression but didn't alter the significance of variables).

While I didn't pull in the population data (that's a good idea), I added the following independent variables:

--year over year % change in GDP,
--a dummy variable for films released during the Cold War years,
--a dummy variable for, sci fi/fantasy/special effects-driven films (by my reckoning, you could say 2001: A Space Odyssey was one of the first such film...it's a judgement call by yours truly if a movie fit this definition so any errors would be mine),
--a dummy variable for cartoons,
--a dummy variable for films released after the original Star Wars was released (note this is also effectively a cable TV variable), and finally
--an index reflecting the number of years before and after the original Star Wars was released (note this is also likely picks up the effects of improving technology and, to some effect, the advent of a cable TV variable) .

As noted, I used inflation adjusted gross receipts as the dependent variable. I also modeled the ranking of the films (i.e. Gone with the Wind took on a value of 1, etc.) which generally had little impact.

I won't post the results (this isn't The Journal of Econometrics) but basically:

-the regressions were significant but r-square was only about 0.2,
-the ONLY significant variables (p-value less than 0.01) were the dummy for sci fi/fantasy/special effects-driven films AND the Star Wars index variable.
--the coefficient on the sci fi/fantasy/special effects-driven films dummy was $95 million, which (for the sake of simplicity well say) means such films make $95 million than other such films
--however, the NEGATIVE $5 million coefficient on the Star Wars index variable suggests that the avg inflation adjusted gross revenue has DECLINED $5 million every year since the original Star Wars was released.
-- the cartoon dummy was only significant (p-value of 0.08) in the model using the movie ranking,

In general, these results support LS's contention that Star Wars hasn't ruined films, and that sci fi etc films are likely more popular because of the absence of meaningful substitutes. However, the coefficients also support a view that the sci fi etc effect of $95 million (if it was spawned by Star Wars) was wiped out 19 years after 1977 (i.e. The negative $5 million coefficient * 19).