RDP 2025-01: Are Investment Tax Breaks Effective? Australian Evidence Appendix C: Testing for the Effect of Corporate Tax Changes

Nu Nu Win, Jonathan Hambur and Robert Breunig

February 2025

To explore the possible effect of the corporate tax cuts, we re-estimate the DD models using buildings & structures investment from the CAPEX data described above. This asset class is ineligible for the investment incentives that we study in this paper.

Looking at the results for the GFC policy period, we can see there is no significant increase in investment in buildings & structures either on the intensive or extensive margins. There appears to be no complementarity between investment in these asset types (i.e. increases in eligible machinery & equipment have no effect on investment in buildings & structures). This suggests that looking at buildings & structures can be a useful placebo test for other ITB policies.

In contrast, for the 2015, and to a lesser extent 2016, policies there is evidence of an increase in buildings & structures investment, suggesting that firms reacted to the corporate tax cuts.[17] As firms appear to be responding to the corporate tax cuts, we should be cautious about interpreting our findings regarding the effects of the 2015 and 2016 ITB.

Table C1: DD Regression Results for Building & Structures
  Intensive margin   Extensive margin
All firms Companies Unincorporated All firms Companies Unincorporated
GFC 2009 −0.017
(0.440)		 −0.074
(0.526)		     −0.242
(0.173)		 −0.182
(0.172)		 −0.449
(0.285)
Observations 1,250 734     25,880 13,330 12,518
Small 2015 0.665**
(0.312)		 0.829*
(0.448)		     0.010
(0.226)		 −0.157
(0.269)		 0.070
(0.333)
Observations 1,317 761     16,808 9,040 4,008
Medium 2016 0.171
(0.323)		 0.294
(0.351)		     0.219*
(0.125)		 0.156
(0.134)		 0.277
(0.181)
Observations 1,897 1,493     10,612 7,651 2,908

Notes: Results exclude mining. Regressions do not include controls due to small sample size. ***, ** and * denote statistical significance at the 1, 5 and 10 per cent levels, respectively. Cluster robust standard errors in parentheses.

Footnote

Note that intensive margin unincorporated results are not shown due to the very small sample of unincorporated small firms doing building investment. [17]