Statistics in the News: Chapter 18 Model Building With Multiple Regression

Every February, members of the Academy of Motion Picture Arts and Sciences vote to nominate five films for an Academy Award (an Oscar). Then, six weeks later after a second round of voting, winners in over 20 major categories are announced at a nationally televised ceremony. For actors, actresses, directors, and producers a nomination or award can translate into a great deal of money. So one may well ask: What is an Oscar worth?

Recently, four researchers tackled the question by building two alternative econometric models with the help of weekly box-office data for 131 nominated and 131 non-nominated films that were released during the 1978-1987 period. One equation in each model related SHARE (the percentage of all screens on which a given film appeared in a given week) to a number of dummy variables (including film-specific factors and the quarter of the year in which the film was released), the number of weeks the film remained in distribution, and variables measuring the impact of nominations or Academy awards for best supporting actor/actress, best actor/actress, and best picture. A second equation in each model related ARPS (the average revenue per screen) to the same variables.

Table A excerpts some of the results derived from one of the models. The separate effects on weekly revenues of a best actor/actress nomination or award appear in the top half of the table. Corresponding data for a best picture nomination or award appear in the lower half. The percentages indicate the release dates of the affected films.

Table A Predicted Weekly Values of an Oscar Nomination and Award (1997 dollars)

Thus, 13% of all films with best actor/actress nominations were released in the 1^st quarter of the year and such a nomination raised weekly revenues for such a film by $561,733. On the other hand, 61% of all films with best actor/actress nominations were released in the 4^th quarter of the year and such a nomination raised weekly revenues of such a film by $3,753,359. And note: In each case, the revenue figure represents additional revenue compared to the most successful non-nominated film. Hence the dollar amount might be viewed as the minimum value of an Oscar. The remaining entries are similarly interpreted. For example, a 4^th quarter film receiving an Oscar for Best Picture can be expected to garner an additional $17,879,921 per week.

The authors explain the greater payoff to late-release films in two ways:

• by the fact that 4^th quarter films are apt to be fresh in the minds of voting members of the Academy and, therefore, have a greater chance of being nominated and of later receiving an Oscar than, say, 1^st quarter films,
• by the fact that 4^th quarter films have a greater chance to be still playing in theaters when the February nominations are made, which in turn gives them a greater chance than, say, the previous year's 1^st quarter releases to be still playing at awards time and in the following weeks.

Sources. Adapted from Randy A. Nelson, Michael R. Donihue, Donald M. Waldman, and Calbraith Wheaton, "What's An Oscar Worth?" Economic Inquiry, January 2001, pp. 1-16.