Estimates of government spending multipliers in Australian data

In this paper, cumulative multipliers for government consumption spending and for government investment spending are estimated by the method of local projections with instrumental variables using quarterly Australian data from 1984Q1 to 2019Q4. This method directly estimates dollar-for-dollar cumulative multipliers. The estimates are for ‘average’ or non-state-dependent multipliers, and for multipliers which are state dependent. Government consumption spending multipliers have been estimated for Australia but not conditional on an economic state and government investment spending multipliers have not received attention. In this paper, the state-dependent multipliers are estimated for states of high and low unemployment. The peak cumulative multiplier for government investment spending is −0.95 in the non-state-dependent case and −2.54 in the low unemployment state, both for a horizon of six quarters. Government investment spending crowds out private sector activity and has a negligible effect in the high unemployment state. Government consumption spending crowds in (expands) private sector activity only in the high unemployment state and only moderately and has a negligible effect otherwise. The cumulative government consumption spending multiplier in the high unemployment state is 0.94 at six quarters. The results are robust to variations in the baseline specification.