Using VAR analysis on US data, we show that unanticipated fiscal expansions boost private consumption and business formation. Models with an extensive investment margin, i.e. endogenous firm and product entry, have difficulties explaining these two phenomena simultaneously. Considering different variants of an endogenous-entry business cycle model, we show that crowding-in of both consumption and entry can be generated only under very specific assumptions. In a static model with full depreciation, labor supply has to be extremely elastic. In a dynamic model, the fiscal stimulus must be sufficiently persistent such that future profits are high enough to generate entry. However, consumption falls for conventional parameter values. Lowering the wealth effect through the introduction of rule-of-thumb consumers or GHH preferences does not help to bring the model closer to the data.