A comment on evaluating the cost-effectiveness of armored tactical wheeled vehicles

This comment discusses the pros and cons of the methodology and data used in our previous study on the cost-effectiveness of armor on Tactical Wheeled Vehicles (TWVs), and responds to recent critiques by Franz Gayl. In our previous article, we evaluated the large-scale Army policies to replace relatively light Type 1 Tactical Wheeled Vehicles (TWVs) with moderately protected Type 2 variants, and later to replace Type 2s with heavily protected Type 3s. We find that the switch from Type 2 to Type 3 TWVs did not appreciably reduce fatalities and were not cost-effective. Mr. Gayl contends that the data and choice of control variables used in our original study negatively bias our findings for Type 3 TWVs. We defend our previous conclusions and argue that Gayl's suggested approach of focusing on <italic>deaths per insurgent attack</italic> fails to account for effects of the vehicles on when, where, and how attacks occurred. Our methodology does not suffer from this bias and measures effects on total unit casualties rather those incurred <italic>per attack</italic>. We explain that our estimates are stable across many specifications and are not sensitive to the choice of controls as Gayl suggests.