We use new data on cyclically adjusted primary balances for Latin America and the Caribbean to estimate effects of fiscal consolidations on GDP and some of its components. Identification is conducted through a doubly-robust estimation procedure that controls for non-randomness in the ”treatment assignment” by inverse probability weighting and impulse responses are generated by local projections. Results suggest output contraction by more than one percent on impact, with economy starting to recover from the second year on. Composition effects indicate that revenuebased adjustments are way more contractionary than expenditure-based ones. Disentangling effects between demand components, we find consumption being in general less responsive to consolidations than investment, although nonlinearities associated to initial levels of debt and taxation might play an important role.

ACCESS PAPER