Conventional wisdom suggests that macroeconomic outcomes do not follow a political business cycle (PBC) pattern. In this study, the nature of the electoral underpinnings of such opportunistic behaviour are investigated by analysing alternative formulations of PBC theory: (1) a naïve-unconditional PBC, (2) an electoral security-conditional PBC, (3) an electoral uncertainty-conditional PBC; and (4) a partisan-conditional PBC. Data on the US real personal income growth rate for the 1948:1–2000:4 quarterly period reveals support for both naïve-unconditional and partisan-conditional PBCs, yet rejects an electoral cycle attributable to the incumbent administration's ex ante re-election prospects. Simulation analysis reveals that while Democratic administrations enjoy higher income growth than Republican counterparts for non pre-election stimulus periods, Republican presidents are associated with larger pre-election economic expansions than Democratic presidents consistent with a partisan-conditional PBC theoretical model. This finding supports the notion that incumbent governments engage in partisan-based policy balancing with respect to the creation of opportunistic electoral cycles in real macroeconomic activity. On a broader level, these statistical findings provide strong support for adaptive models of the electoral cycle that emphasize partisan differences while refuting the policy neutrality proposition and rational-competence models predicated on the extent to which an incumbent administration experiences electoral vulnerability.