Inflation in weekend box office estimates

Abstract

Do movie studios inflate box office estimates that they report every Sunday? Is the pattern of this inflation consistent with the studios’ strategic incentives? Analysing actual and estimated box office returns of major movie releases between 2003 and 2010, we find that movie studios substantially overestimate weekend box office performance and that this overestimation is highly unlikely to be due to chance. Our findings also suggest that box office inflation is strategic because (1) inflation is substantially higher in the first weekend of release when the incentives are greatest to generate positive word of mouth and (2) inflation for the top two films in a weekend increases with the degree of competition between the films.

Keywords:

• box office
• misreporting
• bias
• marketing

JEL Classification:

• M39
• D21

Acknowledgements

We thank Andrew Healy, David Rothschild and Marc Meredith for valuable suggestions.

Notes

¹ Representative studies include Smith and Smith (1986), Eliashberg and Shuggan (1997), Walls (2005), Einav (2007) and McKenzie (2010). A comprehensive survey can be found in Eliashberg et al. (2006).

² Representative studies include Beneish (1999), Crocker and Slemrod (2007), Degeorge et al. (1999), Burgstahler and Dichev (1997) and Roychowdhury (2006).

³ http://www.boxofficemojo.com.

⁴ http://www.the-numbers.com.

⁵ Average inflation calculated as mean of inflation of all films in sample. For completeness, tests were also run weighting by total revenue. Although the average level of bias is smaller – for example, 3.47% weighted opening week inflation compared with 6.38% – it is still positive and statistically distinguishable from 0.

⁶ Films are classified by the distributing studio. In cases of multiple studios being listed for a film, the analysis uses only the main studio, as listed by Boxofficemojo.com. Manual adjustments were also made to account for studio parent ownership and changes in ownership structure.

⁷ Although variance may decline due to any improvement in accuracy over the lifetime of a release, we see no fundamental reason why the bias should decline.

⁸ In 2010 ticket-price inflation adjusted dollars (Box Office Mojo, 2011).