Concerns over free riding are widespread in alliances, in particular NATO. An intuitive approach to analyzing free riding is treating it as spatial interdependence between the allies: How does a country’s defence spending react to changes in its allies’ military expenditures? Recent work has found a significant negative response for NATO allies and implied that the alliance suffers from substantial free riding. However, this research does not quantify the concerned effect. Furthermore, it does not adequately account for temporal dependence, which risks bias in the parameter of spatial interdependence. Directly accounting for temporal dependence provides a meaningfully different picture of the within-alliance dynamics, revealing that the long-run effect of free riding is indeed more significant than its short-run effect, as suggested by static spatial (autoregressive) models. We discuss the relevant practical and theoretical implications.