Academic journal article Economic Inquiry

Market Entry Mode: Evidence from the Golden Age of Hollywood

Academic journal article Economic Inquiry

Market Entry Mode: Evidence from the Golden Age of Hollywood

Article excerpt


A firm may use different modes of entering a market. How firms decide what mode to use is important because the cost structures across modes are different. Therefore, exports will respond to shocks differently depending on the mode used (Bernard. Grazzi, and Tornasi 2015). However, data constraints have restricted our knowledge of this decision. For example, Arkolakis, Ramondo, Rodriguez-Clare, and Yeaple (2013) do not consider licensing since "there is little data" on it. Firm-level trade data sets generally do not link a transaction to a firm unless the firm exports directly. (1)

This article examines market entry modal choice for U.S. motion pictures in their earliest years. It uses novel data sources that provide very detailed information on market entry. The Film Daily Year Book reports the locations of sales offices for U.S. studios. I examine the period 1921-1935, the period of U.S. studios' initial expansion overseas. I augment these data with internal company data for United Artists (UA) from 1928 to 1949. These data give country-level detail on mode of entry, sales, and distribution costs. In addition to a rich set of data, the archives include memos and other documents that give direct insight into how the decisions to enter markets were made.

These data sources, which have not been used by economists before, improve on the available data in two major ways. This industry and time period give us an unusually clear view of how companies build out their international distribution networks, providing insight into the forces of gravity. The data begin with the first year that U.S. studios established overseas sales offices. Motion pictures were a new technology and the companies that produced them were either new entrants or firms that produced nontraded entertainment, vaudeville theaters.

They also allow us to examine the frictions that determine modal choice. While the theoretical literature has emphasized contracting frictions in foreign direct investment (FDI), the empirical literature is sparse. Existing FDI data do not give information about the revenues of licensees becaue they are unaffiliated foreign companies and licensing agreements are typically confidential. The UA data give us rare access to compare total revenues generated across modes, not just revenue to the licensing studio.

As an analytical framework, I set out a model adapted from Helpman, Melitz, and Yeaple (2004) and Yeaple (2009) to reflect the modes that movie studios use to distribute their films. Modes are defined by the share of the distribution cost that the firm pays. Using a more intensive mode--paying more of the distribution cost--allows the firm to keep a greater share of the revenue. The most intensive mode, opening a foreign office, requires the firm to pay all the distribution costs but allows it to keep all the revenue. Studios could pay less of the distribution cost by using licensed agents or an even less cost intensive export sale, but would receive a lower share of revenue. (2)

I develop a number of empirical facts that are consistent with the model. Studios use a hierarchy of modes to export, using modes where it pays more of the costs in larger markets. There is a strong relationship between market size and the number of U.S. major studios with an office. However, there is a small role for distance in choosing where to site offices. Companies set up offices in distant large markets early in the process of building out their network. These served as platforms to expand out. These data indicate that directed search is important and can explain why trade is less geographically concentrated than predicted by many modern models of trade. It also suggests that headquarters have relatively little interaction with affiliates, as emphasized by Arkolakis et al. (2013), Atalay, Hortacsu, and Syverson (2014), Ramondo (2014), and Ramondo, Rappoport, and Ruhl (2015). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.