Additional Concerns About O'Brien and Lane's Article on Film Incentive Programs

In this note, we follow up on the 2020 exchange in this journal between John Charles Bradbury and the replying authors Nina F. O’Brien, and Christianne J. Lane. We focus on some inconsistencies in O’Brien and Lane’s work, particularly regarding a critical maximum value in their dataset, the significance of which is reinforced by some publicly available data that is new to this discussion. In examining the economic impact of film incentive programs, O’Brien and Lane use the independent variable Incentives(millions) to measure “the allocation a state’s legislature earmarks for a programme.” However, some states’ programs have uncapped incentives; for these states, O’Brien and Lane proceed to estimate the ‘size’ of these ‘allocations’ on the basis of the highest allocation in any state with capped incentives. The result is an independent variable that is a mix of actual allocations for states with legislatively capped allocations and estimated allocations for states without legislatively capped allocations. Here we find that the aforementioned maximum value of Incentives(millions) is likely inaccurate, and that the total value of allocations estimated by O’Brien and Lane on the basis of that value likely greatly exceeds the value of actual allocations, casting their results in further doubt. Perhaps some of that doubt could be mitigated by the public release of their dataset, were that to occur.