The Welfare Effects of Chinese Industrial Inequality Policy: The Case of Movie Cinemas

Abstract

This study explores China's movie theater industry and quantifies the effects of a place-based industrial policy on cinema entries. Particularly, it aims to discuss the policy effect in the context of regional inequalities. The analysis consists of three parts. In the first part, titled "Demand for Watching Movies in China", I estimate a nested logit model with instrumental variables for the price to analyze Chinese consumers' demand for movies. The instrumental variables are the average attributes of the rival cinemas in the same market. This model identifies consumers' willingness to pay for movies, which is critical to the later welfare gains calculation. Further, the demand model finds that China's movie-goers favor premium screens and the current industry-wide upgrading for giant screens is in line with consumers' tastes. The second part, titled "The Effect of the Build-and-subsidize Policy on Cinema Entries", estimates the effect of the build-and-subsidize policy on the number of cinema entries. The build-and-subsidize policy offers one-time subsidies to new cinemas in county-level cities and counties in central and western China, and less developed counties in eastern China. These subsidies can cover roughly 60% of construction costs for a typical eligible cinema. By leveraging the difference-in-difference model, I find that the build-and-subsidize policy increases the number of new cinemas by 0.48 per county per year during the policy period. Although there is a reversal right after the policy concludes, the net effects are still positive (0.54 cinema per county). This estimated model is another essential component for calculating welfare gains: it would be used to generate counterfactual sets of cinemas with and without subsidies. The final part, titled "Consumers’ Welfare Gains from the Build-and-subsidize Policy", combines the estimated models in the first two parts to calculate consumers' welfare gains. I first use the estimated difference-in-difference model to predict the new cinema counts with and without the subsidies. Based on the predictions, I construct consumers' choice sets under the two scenarios and use the estimated logit model to calculate the change in consumer surplus. The calculation quantifies residents' benefits from the policy. I find that a typical resident in the subsidized areas receives 0.58 RMB welfare gains per year from the enlarged choice set caused by the build-and-subsidize policy. Overall, it takes 4.9 years to recover the subsidies. Further, there is large heterogeneity in policy effectiveness across areas. Wealthier and more populous areas are associated with larger welfare gains from the subsidies. Although the policy has mitigated the cinema distribution inequality between the coastal provinces and the inland provinces, it has not narrowed the gap within the subsidized areas.