Historical Analysis of Business Investment and Taxes Going Back to 1961
Here is a link to the CCPA study we released yesterday, analyzing the determinants of business fixed non-residential capital investment spending in Canada on the basis of quarterly data from 1961 through 2010.Â It formally tests for the direct significance of corporate tax variables and finds no such evidence (in either univariate or multivariate analysis).Â There is still an indirect impact of tax rates on investment experienced via incremental corporate cash flow, but that indirect effect is weak and getting weaker.Â In recent years, after controlling for other determinants (especially economic growth, real interest rates, and oil prices), each dollar of incremental cash flow (via tax cuts or any other mechanism) generates just 10 cents of new investment.
We’d be much better off to simply spend the whole $6 billion on public investment, which would (via GDP effects) “crowd in” almost as much new private investment as would be stimulated with the $6 billion in tax cuts.
There is an appendix dealing with the measurement issues that Stephen Gordon and others used to attack that Globe and Mail study last week. Those arguments would not really apply to this study anyway, since I look at all fixed non-residential capital, not just machinery & equipment, and for that expenditure the pricae deflator has not been declining (although it has increased more slowly than for all GDP). Nevertheless, I argue in the Appendix that the fact that something is becoming relatively cheaper (as computeres are, but no other component of investment), hardly means that people are somehow spending MORE on it — especially as a share of their spendable funds (which cannot be deflated in that manner).Â Dividing deflated investment by deflated GDP is bogus, since they are calculated with completely different deflators — that ratio is not economically meaningful.
Here’s a Toronto Star story which covered the study very well:
There’s a story in Le Devoir, too — suppplemented by a great one-two punch from our own Eric Pineault on similar issues: