Screen Survival of Movies at Competitive Theaters: Vertical and Horizontal Integration in a Spatially Differentiated Market

Abstract

This article empirically examines the effects of vertical integration and horizontal control on the exhibition of films at competing theaters in a market in which some theater owners are integrated into film distribution and others are not. It analyzes the duration of the exhibition runs of movies released by distributors who owned no theaters in Singapore during 2002 and 2003 through a survival model. The estimation shows that films released by these distributors are exhibited for shorter periods in distributor-owned theaters than in independently owned (i.e., unintegrated) theaters, and in theaters that belong to large theater chains than in the theaters of small theater chains. A spatial aspect of competition among theaters with different owners is revealed in the finding that a film is more likely to stay on at the theater at a given time when an adjacent rival theater still runs the same title. The increase in the probability of screen survival due to the spatial rivalry is greater the nearer the competing theater. These results are found when both film and theater-specific variations are controlled for.

Notes

¹ Ford and Jackson (1997) measured a net reduction in social welfare owing to the integration of cable systems and program channels, indicating that the exclusionary effect outweighs the efficiency effect. Chipty (2001) also detected the foreclosure of rival channels by integrated operators, but calculated that the exclusionary loss is slightly offset by the efficiency gains. Waterman and Weiss (1996) found that integrated cable systems favor owned networks in terms of carriage and marketing treatment, but did not determine which effect is the dominant cause. The study of Oba and Chan-Olmsted (2006) about the programming choices of TV stations ascribed the evidence to reasons of efficiency rather than foreclosure.

²Some studies, however, looked at the effect of horizontal concentration in the U.S. cable system market, where the firms are local monopolies. Chipty (1995) indicated that the size of a Multiple Cable System Operator (MSO) increases its bargaining power relative to network suppliers. However, Chipty and Snyder (1999), controlling for the efficiency dimension, did not detect such a result. Further, Waterman and Weiss (1997) argued that horizontal market control makes it easier to realize the exclusionary potential of vertical integration.

³See De Vany and Walls (1997) for a rigorous description of movie distribution contracts.

⁴The other efficiency advantages of vertical integration in the movie business are specifically discussed. Integration can tame the uncertainty and volatility of audience demand, direct sales revenue into production, reduce transaction costs in distribution and marketing (Blackstone & Bowman, 1999; De Vany, 2004; Litman, 1998; Waterman & Weiss, 1997), and mediate marketing incentive externalities (De Vany & Eckert, 1991; Hanssen 2000, 2002; Kenney & Klein, 1983, 2000). These factors also can drive integrated exhibitors to favor their own films over others.

⁵Singapore has no general anticompetition legislation or enforcement, although very recently the government has prepared such in the wake of its free trade agreements. In 2003, the “Media Competition” code was promulgated by the media regulation authority under the 2002 Media Development Act, but the code's jurisdiction covers only television, radio, and newspapers, not cinemas.

⁶There are two other independent theaters that feature only ethnic films, such as Indian (Bollywood) and Malay productions. These theaters are not included in this study.

⁷Gil (2007, 2008) and Agostini and Saavedra (2006) characterized the distribution–exhibition integration in other cinema industries in the same way. This current context of vertical integration in the movie industry contrasts with the American situation, where production studios were integrated with distributors and sometimes exhibitors.

⁸Open bidding by exhibitors for the right to show a title is not in place in Singapore.

⁹The model is related to those of Agostini and Saavedra (2006) and Chisholm and Norman (2006), who also analyzed films' theatrical duration, but constructed different factors and did not examine the exhibitions of unintegrated films in focus.

¹⁰The choice of time gap for defining IntgrtExhbtr_ƒt would affect only the coding concerning the Cathay theaters, not those of GV or Shaw, as Cathay has intermittent distributions. Besides, even when the time gap changes to either 1 or 3 weeks, the basic results of the analysis remain.

¹¹All Singapore cineplexes, built mostly during the late 1980s and early 1990s, reside inside major shopping malls, which are at the heart of the town centers or the most populated districts, rather than being free-standing “theater buildings” as in the United States. Such locations are mainly predetermined.

¹²It is an established view in the media economics literature that the consumption quality of media content products, as public goods, hinges directly on content production investments, whose level is driven by the sheer prospects of sales revenue (Fu & Lee, 2008; Hoskins, Mirus, & Rozeboom, 1988; Waterman, 1988; Wildman & Siwek, 1988). Thus, the realized sales receipts (HomeBO_ƒ ) from a home country should closely reflect a movie's value and popularity in another market.

¹³No box-office sales revenue information was available other than the weekly top-10 seller rankings.

¹⁴During the same period, the integrated distributor Shaw distributed 49 films, Golden Village 55, and Cathay 17.

¹⁵Home box-office sales information was not found for 14 films in the sample; their HomeBO_ƒ values were entered with the HomeBO_ƒ average among the other movies from the same country.

^a The standard error for the γ estimate.

*p < .05.

**p < .01.

***p < .001.