Free-Riding and Collective Action in NATO Burden Sharing: Testing the Exploitation Hypothesis Across Seven Decades, 1949–2026

This dissertation asks a question that has shadowed the North Atlantic Treaty Organization since its founding: do the alliance’s wealthiest members shoulder a disproportionate share of the common defence while smaller members free-ride on the protection they receive? The exploitation hypothesis, first advanced by Olson and Zeckhauser (1966), predicts precisely this pattern and has become one of the most cited—and most contested—propositions in defence economics. Against it stands the joint product model of Sandler and Forbes (1980) and Murdoch and Sandler (2000), which holds that defence yields not only a pure, non-excludable deterrent but also ally-specific private benefits that blunt the incentive to free-ride. This study adjudicates between the two using a time-series cross-sectional panel of all 32 NATO member states observed across the alliance’s full lifespan, 1949 to 2026, assembled from NATO, SIPRI, World Bank, IMF, and Correlates of War sources. The empirical strategy integrates two-way fixed effects, system generalized method of moments estimation, Bai–Perron structural break detection, panel Granger-causality testing, and spatial econometrics. Four findings emerge. First, exploitation is era-contingent rather than universal: it is strongest under the doctrine of Massive Retaliation, reverses after September 11 as frontline states outspend their larger partners, and returns partially after 2022. Second, structural breaks cluster at 1958, 1968, and 1993, aligning with doctrinal watersheds, while the exploitation coefficient declines quasi-monotonically. Third, spending decisions are strategically interdependent, with causality running from the largest members to smaller allies until a post-2014 reversal, and a pronounced threat gradient tied to proximity to Moscow but no measurable spatial diffusion. Fourth, burden-sharing is deeply path dependent. Taken together, the evidence supports the joint product interpretation: the publicness ratio—the balance between pure public and ally-specific benefits—governs the strength of free-riding, and the alliance has migrated from a pure public good provider toward a mixed club. The study offers the most temporally complete test of the exploitation hypothesis to date and an evidence base for the 2% and 2025 Hague 5% frameworks.