My JLF colleague Jon Sanders has a column out on film subsidies that I recommend you read. It}s a well-crafted look at film subsidies and the economic impact studies used to justify them. A brief summary:

Industry studies, meanwhile, are notorious for promoting the idea of a greater return on investment, mainly by ignoring opportunity costs and assuming that any job in the industry exists because of the subsidies. Even so, the figures reported by MPAA are rather sensationalized. A close reading of the report methodology, which used the IMPLAN input-output model, clues us in (emphasis added):

This film, like other productions, has an amount of direct expenditure, such as equipment rentals, which in turn stimulates so-called indirect impacts across the supply chain of the equipment rental industry. These expenditures also create new jobs as firms add labor to meet rising demand, which in turn raises incomes and stimulates what are known as induced impacts as higher incomes flow through to consumption. The total economic impact of the production expenditures is the sum of each of these direct, indirect, and induced impacts.

In short, this is a model of sums — it is all plus signs. Despite what it says, it doesn’t measure the total economic impact; it measures only benefits expanded by whatever multipliers it used (more on that below). It doesn’t take costs into account.