Phased-In Tax Cuts and Economic Activity

This paper by Christopher House and Matthew Shapiro (AER 2006; NBER WP 10415, April 2004) uses a dynamic general equilibrium model to quantify how the timing of a legislated tax-rate change – not just its eventual size – affects employment, output and investment. The central mechanism is intertemporal substitution: when a labor-income tax cut is phased in rather than immediate, workers and firms have an incentive to defer work and investment until the lower rate actually takes effect, so in the model “there is a substantial reduction in economic activity during the period of the delay” even though the tax cut eventually raises long-run output. Capital-income tax cuts behave in the opposite way: delaying one can increase current investment, because firms rush to invest while depreciation write-offs are still valued against the current (higher) tax rate, before the cut reduces that benefit. Applying the model to the actual 2001 and 2003 U.S. tax laws, the authors argue the 2001 law’s multi-year phase-in schedule plausibly slowed the 2001-2003 recovery, while the 2003 law’s acceleration of those same cuts to take effect immediately plausibly contributed to the growth pickup in the second half of 2003; they further argue that federal budget-scoring rules (the ten-year revenue window and the Byrd rule’s supermajority requirement for provisions beyond it) are themselves a structural reason legislators keep choosing phased-in and sunsetting tax changes. These are model-based, calibrated simulation results, not causal estimates from a natural experiment, and the authors are explicit that other factors also affected the 2001-2003 period.

Christopher L. House | Macro Paper Warehouse

Phased-In Tax Cuts and Economic Activity