The latest edition of the Canadian Tax Journal [gated] has a nice article by Daria Crisan and Kenneth J McKenzie that documents Canada's relatively generous tax subsidies for R&D spending, yet relatively underwhelming investment by large corporations, over the period 1981-2016. The article briefly summarizes the chronology of federal R&D programs and gives an overview of provincial policies. It finds that while Canada's spending on R&D is high relative to peer countries, the amount of R&D being undertaken relative to GDP consistently underperforms these peers and continues to decline. The bulk of the article is a presentation of data showing these trends. The conclusion is intriguing, positing three possible explanations for the puzzle of high spending but low investment:
The first is the rather obvious point that it may well be that Canada’s r&d performance would have been even worse in the absence of the subsidies. Of course we don’t observe this counterfactual, but it is consistent with the above observations.
The second comment is more speculative ... Canada relies much more than
other countries on the type of “indirect” tax subsidies that we consider here, which
are generally available to all companies, as opposed to “direct” subsidies, such as
targeted grants. It could be that the nature of r&d subsidies in Canada—the reliance
on indirect tax incentives rather than direct grants—is the problem. ...
This leads to our third, and final, observation. To our knowledge, there is in factThe authors conclude that their own ongoing research involves an empirical investigation of the effectiveness of direct and indirect incentives in promoting business r&d investment in Canadian
very little rigorous empirical evidence regarding the efficacy of direct versus
indirect government subsidies for r&d. Moving in this direction may well be the
right thing to do, but this seems to us to be based more on faith, and perhaps some
frustration with the Canadian “r&d policy puzzle,” than on solid empirical evidence.
Our hope is that the data presented here provide, at least in part, the basis
for additional research in this regard in a Canadian context.
I don't know whether the type of spending matters in terms of investment incentives. I would think that it matters what the spending is related to. For example, does spending a lot of money on companies to patent things actually lead to "innovation," whatever that word means? What I have read to date suggests not. There seems to be a strong connection of innovation spending toward traditional legal rights in copyright and patent, but these rights seem decreasingly relevant to many contemporary innovative business models.
Despite a general lack of empirical evidence that taxpayer dollars are well spent on R&D subsidies, governments everywhere spend and spend and spend to spur innovation. As a result empirical studies that shed light on the efficacy of this spending will always be welcome. If the studies show that traditional modes of subsidizing R&D do not provide the intended results, the question is whether governments themselves will be willing and able to innovate in terms of how they support innovation. From the chronology presented herein and my own research on the topic, the prospects seem dim. I look forward to seeing more of this important research.