Scholars systematically mismeasure power resources and military burdens by using gross domestic product (GDP) as a proxy for the income states can devote to arming. The core problem is that GDP confounds two conceptually distinct forms of income into one additive indicator. Subsistence income represents resources needed to provide the “bread” necessary to cover the basic subsistence needs of the population. Surplus income represents the remaining resources that could be allocated to “guns” or “butter.” Our new measure of surplus domestic product (SDP) corrects for this measurement error by decomposing subsistence income and surplus income from total GDP. Validation exercises demonstrate that SDP outperforms GDP at measuring the distribution of power resources. Though theoretically we expect states’ decisions to arm are influenced by the distribution of power; empirical models using GDP find mixed support for this expectation. Strikingly, using SDP reveals strong support for this proposition.