Zwick & Mahon (2017): Tax Policy and Heterogeneous Investment Behavior
AER quasi-experimental study of US bonus depreciation across 120,000+ firms: immediate-expensing provisions raised eligible investment by roughly 10–17%, with far larger responses among small and cash-constrained firms — the cleanest identification in the whole rent-targeting tax file that moving th
Summary
Zwick and Mahon, "Tax Policy and Heterogeneous Investment Behavior" (American Economic Review 107(1), 2017, 217–248), study US bonus depreciation — temporary partial immediate expensing of equipment investment — using tax-return data on over 120,000 firms and a difference-in-differences design comparing industries whose equipment mix made them more or less eligible.[1]
Findings
- Bonus depreciation "raised investment in eligible capital relative to ineligible capital by 10.4% between 2001 and 2004 and 16.9% between 2008 and 2010" (abstract; verified against the NBER working-paper text this session).
- Responses were dramatically heterogeneous: "small firms respond 95% more than big firms," and firms "respond strongly when the policy generates immediate cash flows but not when cash flows only come in the future" — the authors read this as evidence that "financial frictions or fixed costs amplify investment responses," not just the cost of capital.[1]
Why This Matters for Rent-Targeting Tax Design
Immediate expensing is the defining feature of the cash-flow tax — it is what removes the normal return from the base, leaving only economic rent. Zwick–Mahon is the cleanest quasi-experimental evidence that moving the corporate base toward cash flow stimulates real investment: the mirror image of the distortion a standard income-type base imposes. The TCJA-era follow-ups by overlapping authors (Chodorow-Reich, Smith, Zidar & Zwick 2024) reach the same design conclusion — accelerated depreciation generated more investment per dollar of revenue than rate cuts.[2]
Within the WS-TECH-RENTS file this is the positive investment evidence that sits alongside the mixed ACE record (Hebous & Ruf null for multinationals' production investment; Konings et al. positive for Belgian affiliates): expensing-type rent exemption has the strongest identified investment effects, while allowance-type designs show clearer financing effects than real ones. The synthesis outcome must keep the two instrument families distinct.
See Also
- Cash-Flow Tax — the design this evidence supports
- Allowance for Corporate Equity — the accrual sibling with a more mixed record
- Hebous & Ruf (2017) — the contrast case
- Objection: Taxing quasi-rents kills innovation
Sources
- Eric Zwick & James Mahon (2017), "Tax Policy and Heterogeneous Investment Behavior," American Economic Review 107(1), 217–248 (working-paper version: NBER WP 21876) — used for the investment-response magnitudes and heterogeneity findings (abstract quoted verbatim; verified against the NBER PDF this session). AEA · NBER WP 21876
- Gabriel Chodorow-Reich, Matthew Smith, Owen Zidar & Eric Zwick (2024), "Tax Policy and Investment in a Global Economy" (NBER WP 32180) and "Lessons from the Biggest Business Tax Cut in US History," Journal of Economic Perspectives 38(3), 61–88 — used for the TCJA-era corroboration of expensing's cost-effectiveness. NBER · AEA